Refine Search

New Search

Results: 50

(searched for: 10.29328/journal.aaai.1001011)
Save to Scifeed
Page of 1
Articles per Page
by
Show export options
  Select all
Alan M. Thompson
Published: 1 July 1980
Journal: AI Magazine
AI Magazine, Volume 1, pp 1-1; https://doi.org/10.1609/aimag.v1i1.83

Abstract:
As a major scientific society, the AAAI has a responsibility for promoting its field as well as informing its members of the latest technical developments. Since the latter function is adequately performed by the several journals and conference proceedings already mentioned, the editorial committee chose to assign to AI Magazine the task of providing AAAI members and the public as well with a broader perspective on the research activities within AI. The approach we intend to take includes publishing informative expository and survey articles designed not so much for those working within a particular problem domain, but for those outside it who would like to gain a better understanding of the issues and methods currently being studied without having to cull all the technical literature.
Xiangyu Zhao
ACM SIGWEB Newsletter pp 1-4; https://doi.org/10.1145/3533274.3533277

Abstract:
Dr. Xiangyu Zhao is an assistant professor of the school of data science at City University of Hong Kong (CityU). Prior to CityU, he completed his PhD (2021) at MSU under the advisory of Dr. Jiliang Tang, MS (2017) at USTC and BEng (2014) at UESTC. His current research interests include data mining and machine learning, especially (1) Personalization, Recommender System, Online Advertising, Search Engine, and Information Retrieval; (2) Urban Computing, Smart City, and GeoAI; (3) Deep Reinforcement Learning, AutoML, and Multimodal ML; and (4) AI for Social Computing, Finance, Education, Ecosystem, and Healthcare. He has published more than 30 papers in top conferences (e.g., KDD, WWW, AAAI, SIGIR, ICDE, CIKM, ICDM, WSDM, RecSys, ICLR) and journals (e.g., TOIS, SIGKDD, SIGWeb, EPL, APS). His research received ICDM'21 Best-ranked Papers, Global Top 100 Chinese New Stars in AI, CCF-Tencent Open Fund, Criteo Research Award, Bytedance Research Award and MSU Dissertation Fellowship. He serves as top data science conference (senior) program committee members and session chairs (e.g., KDD, AAAI, IJCAI, ICML, ICLR, CIKM), and journal reviewers (e.g., TKDE, TKDD, TOIS, CSUR). He serves as the organizers of [email protected]'19, [email protected]'20, 2nd [email protected]'21, 2nd [email protected]'21, and a lead tutor at WWW'21/22 and IJCAI'21. He also serves as the founding academic committee members of MLNLP, the largest AI community in China with 800,000 members/followers. The models and algorithms from his research have been launched in the online system of many companies.
C. Basu, H. Hirsh, W. W. Cohen, C. Nevill-Manning
Journal of Artificial Intelligence Research, Volume 14, pp 231-252; https://doi.org/10.1613/jair.739

Abstract:
The growing need to manage and exploit the proliferation of online data sources is opening up new opportunities for bringing people closer to the resources they need. For instance, consider a recommendation service through which researchers can receive daily pointers to journal papers in their fields of interest. We survey some of the known approaches to the problem of technical paper recommendation and ask how they can be extended to deal with multiple information sources. More specifically, we focus on a variant of this problem -- recommending conference paper submissions to reviewing committee members -- which offers us a testbed to try different approaches. Using WHIRL -- an information integration system -- we are able to implement different recommendation algorithms derived from information retrieval principles. We also use a novel autonomous procedure for gathering reviewer interest information from the Web. We evaluate our approach and compare it to other methods using preference data provided by members of the AAAI-98 conference reviewing committee along with data about the actual submissions.
Claus Atzenbeck
ACM SIGWEB Newsletter pp 1-4; https://doi.org/10.1145/2749279.2749280

Abstract:
Munmun De Choudhury is an assistant professor at the School of Interactive Computing, Georgia Tech and a faculty associate with the Berkman Center for Internet and Society at Harvard. Munmun's research interests are in computational social science, with a specific focus on reasoning about health behaviors from social digital footprints. She has been a recipient of the Grace Hopper Scholarship, recognized with an IBM Emergent Leaders in Multimedia award, and recipient of ACM SIGCHI 2014 best paper award and ACM SIGCHI honorable mention awards in 2012 and 2013. Munmun's work has been extensively covered in popular press as well, including the New York Times, the TIME magazine, the Wall Street Journal, and the NPR. She served as the Program Chair of the International AAAI Conference on Weblogs and Social Media (ICWSM) in 2014, and currently is a member of the Steering Committee of this interdisciplinary conference. Earlier, Munmun was a postdoctoral researcher at Microsoft Research, a research fellow at Rutgers, and obtained a PhD in Computer Science from Arizona State University in 2011.
Parikshit N. Mahalle Aboli H. Patil
Turkish Journal of Computer and Mathematics Education (TURCOMAT), Volume 12, pp 7-15; https://doi.org/10.17762/turcomat.v12i4.455

Abstract:
Peer review is one of the most crucial and important tasks that are associated with academic conferences, journals and grant proposals; and assignment of an appropriate reviewer plays vital role for accurate and fair review process. This paper presents a learning based proactive system that assigns reviewer(s) whose expertise matches with the domain(s) of the paper satisfying constraints. The assignment of reviewer to paper needs to satisfy various constraints such as maximum number of papers per reviewer, minimum number of reviewers per paper and conflict of interest. he core challenge in reviewer paper assignment is to make the computer understand the subject domain of experts and papers. In proposed system, features are extracted from title, abstract and introduction section of publications of reviewer and submitted papers. These features help the model learn the domain features of experts and submitted papers more accurately. Once the training set is built utilizing the inherent correlation between abstract and title, the model is trained and the similarity between reviewers and papers is predicted. The experimental results on test data set of AAAI 2014 and NIPS 2019 demonstrate the effectiveness of the proposed system.
Mohammad T. Irfan, Luis E. Ortiz
Published: 1 October 2014
Artificial Intelligence, Volume 215, pp 79-119; https://doi.org/10.1016/j.artint.2014.06.004

The publisher has not yet granted permission to display this abstract.
Published: 1 January 2010
The publisher has not yet granted permission to display this abstract.
Yun R. Fu
Abstract:
Periodically, the US Army conducts detailed measurement surveys of its soldiers as a way to understand the impact that changes in soldier body size have for the design, fit and sizing of virtually every piece of clothing and equipment that Soldiers wear and use in combat. Recently finished US Army Anthropometric Survey (ANSUR II) has collected 3D body scan data of soldiers at the Natick Solider Center (NSC), as shown in Figure 1. By applying new techniques for shape analysis and classification to these 3D body scan data will help designers of clothing and personal protection equipment to understand and fit Army population. The overall research goal of this proposal is to create a new manifold learning framework for large-scale graph decomposition and approximation problems by low-rank approximation and guarantee computable, stable and fast optimizations for 3D shape description and classification. The PI's group has published (or accepted for publication) 1 book through Springer and 13 scientific papers partially supported by this grant. In particular, these papers are in top journals and conference proceedings such as TPAMI, IJCV, TCSVT, ICCV, AAAI, SDM, ACM MM, etc. One paper, 1 out of 384, receives the Best Paper Award in SDM 2014. The PI, Dr. Y. Raymond Fu has received the 2014 INNS Young Investigator Award, from International Neural Networks Society (INNS), 2014. Leveraged by this grant, the PI has been granted an ARO Young Investigator Program (YIP) Award and a Defense University Research Instrumentation Program (DURIP) award.
Yansong Gao, Jie Zhang
Published: 14 April 2022
Abstract:
The problem of scheduling unrelated machines has been studied since the inception of algorithmic mechanism design~\cite{NR99}. It is a resource allocation problem that entails assigning $m$ tasks to $n$ machines for execution. Machines are regarded as strategic agents who may lie about their execution costs so as to minimize their allocated workload. To address the situation when monetary payment is not an option to compensate the machines' costs, \citeauthor{DBLP:journals/mst/Koutsoupias14} [2014] devised two \textit{truthful} mechanisms, K and P respectively, that achieve an approximation ratio of $\frac{n+1}{2}$ and $n$, for social cost minimization. In addition, no truthful mechanism can achieve an approximation ratio better than $\frac{n+1}{2}$. Hence, mechanism K is optimal. While approximation ratio provides a strong worst-case guarantee, it also limits us to a comprehensive understanding of mechanism performance on various inputs. This paper investigates these two scheduling mechanisms beyond the worst case. We first show that mechanism K achieves a smaller social cost than mechanism P on every input. That is, mechanism K is pointwise better than mechanism P. Next, for each task $j$, when machines' execution costs $t_i^j$ are independent and identically drawn from a task-specific distribution $F^j(t)$, we show that the average-case approximation ratio of mechanism K converges to a constant. This bound is tight for mechanism K. For a better understanding of this distribution dependent constant, on the one hand, we estimate its value by plugging in a few common distributions; on the other, we show that this converging bound improves a known bound \cite{DBLP:conf/aaai/Zhang18} which only captures the single-task setting. Last, we find that the average-case approximation ratio of mechanism P converges to the same constant.
, Lina M Sulieman, Bradley A Malin
Journal of the American Medical Informatics Association, Volume 26, pp 561-576; https://doi.org/10.1093/jamia/ocz009

Abstract:
Objective User-generated content (UGC) in online environments provides opportunities to learn an individual’s health status outside of clinical settings. However, the nature of UGC brings challenges in both data collecting and processing. The purpose of this study is to systematically review the effectiveness of applying machine learning (ML) methodologies to UGC for personal health investigations. Materials and Methods We searched PubMed, Web of Science, IEEE Library, ACM library, AAAI library, and the ACL anthology. We focused on research articles that were published in English and in peer-reviewed journals or conference proceedings between 2010 and 2018. Publications that applied ML to UGC with a focus on personal health were identified for further systematic review. Results We identified 103 eligible studies which we summarized with respect to 5 research categories, 3 data collection strategies, 3 gold standard dataset creation methods, and 4 types of features applied in ML models. Popular off-the-shelf ML models were logistic regression (n = 22), support vector machines (n = 18), naive Bayes (n = 17), ensemble learning (n = 12), and deep learning (n = 11). The most investigated problems were mental health (n = 39) and cancer (n = 15). Common health-related aspects extracted from UGC were treatment experience, sentiments and emotions, coping strategies, and social support. Conclusions The systematic review indicated that ML can be effectively applied to UGC in facilitating the description and inference of personal health. Future research needs to focus on mitigating bias introduced when building study cohorts, creating features from free text, improving clinical creditability of UGC, and model interpretability.
Published: 31 August 2004
The Journal of Allergy and Clinical Immunology, Volume 114; https://doi.org/10.1016/s0091-6749(04)01912-8

The publisher has not yet granted permission to display this abstract.
Published: 30 April 2005
The Journal of Allergy and Clinical Immunology, Volume 115; https://doi.org/10.1016/s0091-6749(05)00436-7

The publisher has not yet granted permission to display this abstract.
Published: 30 September 2004
The Journal of Allergy and Clinical Immunology, Volume 114; https://doi.org/10.1016/s0091-6749(04)02103-7

The publisher has not yet granted permission to display this abstract.
The Journal of Allergy and Clinical Immunology, Volume 113; https://doi.org/10.1016/s0091-6749(04)01274-6

The publisher has not yet granted permission to display this abstract.
Published: 31 December 2003
The Journal of Allergy and Clinical Immunology, Volume 112; https://doi.org/10.1016/s0091-6749(03)02579-x

The publisher has not yet granted permission to display this abstract.
Published: 30 November 2003
The Journal of Allergy and Clinical Immunology, Volume 112; https://doi.org/10.1016/s0091-6749(03)02400-x

The publisher has not yet granted permission to display this abstract.
Published: 30 September 2004
The Journal of Allergy and Clinical Immunology, Volume 114; https://doi.org/10.1016/s0091-6749(04)02106-2

The publisher has not yet granted permission to display this abstract.
Published: 31 March 2005
The Journal of Allergy and Clinical Immunology, Volume 115; https://doi.org/10.1016/s0091-6749(05)00233-2

The publisher has not yet granted permission to display this abstract.
Published: 31 January 2005
The Journal of Allergy and Clinical Immunology, Volume 115; https://doi.org/10.1016/s0091-6749(04)03129-x

The publisher has not yet granted permission to display this abstract.
Published: 30 November 2004
The Journal of Allergy and Clinical Immunology, Volume 114; https://doi.org/10.1016/s0091-6749(04)02500-x

The publisher has not yet granted permission to display this abstract.
Published: 30 June 2005
The Journal of Allergy and Clinical Immunology, Volume 115; https://doi.org/10.1016/s0091-6749(05)01180-2

The publisher has not yet granted permission to display this abstract.
Published: 31 January 2004
The Journal of Allergy and Clinical Immunology, Volume 113; https://doi.org/10.1016/s0091-6749(03)02728-3

The publisher has not yet granted permission to display this abstract.
Published: 29 February 2004
The Journal of Allergy and Clinical Immunology, Volume 113; https://doi.org/10.1016/s0091-6749(04)00584-6

The publisher has not yet granted permission to display this abstract.
The Journal of Allergy and Clinical Immunology, Volume 115; https://doi.org/10.1016/s0091-6749(05)00620-2

The publisher has not yet granted permission to display this abstract.
Published: 28 February 2005
The Journal of Allergy and Clinical Immunology, Volume 115; https://doi.org/10.1016/s0091-6749(05)00039-4

The publisher has not yet granted permission to display this abstract.
Published: 31 March 2004
The Journal of Allergy and Clinical Immunology, Volume 113; https://doi.org/10.1016/s0091-6749(04)00951-0

The publisher has not yet granted permission to display this abstract.
Published: 31 October 2004
The Journal of Allergy and Clinical Immunology, Volume 114; https://doi.org/10.1016/s0091-6749(04)02317-6

The publisher has not yet granted permission to display this abstract.
Published: 30 June 2004
The Journal of Allergy and Clinical Immunology, Volume 113; https://doi.org/10.1016/s0091-6749(04)01444-7

The publisher has not yet granted permission to display this abstract.
Published: 20 May 2019
Journal: Industrial Robot
Industrial Robot, Volume 46, pp 332-336; https://doi.org/10.1108/ir-04-2019-0069

Abstract:
Purpose: The following paper is a “Q&A interview” conducted by Joanne Pransky of Industrial Robot Journal as a method to impart the combined technological, business and personal experience of a prominent, robotic industry PhD and innovator regarding her pioneering efforts and the challenges of bringing a technological invention to market. This paper aims to discuss these issues. Design/methodology/approach: The interviewee is Dr Maja Matarić, Chan Soon-Shiong Distinguished Professor in the Computer Science Department, Neuroscience Program, and the Department of Pediatrics at the University of Southern California, founding director of the USC Robotics and Autonomous Systems Center (RASC), co-director of the USC Robotics Research Lab and Vice Dean for Research in the USC Viterbi School of Engineering. In this interview, Matarić shares her personal and business perspectives on socially assistive robotics. Findings: Matarić received her PhD in Computer Science and Artificial Intelligence from MIT in 1994, MS in Computer Science from MIT in 1990 and BS in Computer Science from the University of Kansas in 1987. Inspired by the vast potential for affordable human-centered technologies, she went on to found and direct the Interaction Lab, initially at Brandeis University and then at the University of Southern California. Her lab works on developing human–robot non-physical interaction algorithms for supporting desirable behavior change; she has worked with a variety of beneficiary user populations, including children with autism, elderly with Alzheimer’s, stroke survivors and teens at risk for Type 2 diabetes, among others. Originality/value: Matarić is a pioneer of the field of socially assistive robotics (SAR) with the goal of improving user health and wellness, communication, learning and autonomy. SAR uses interdisciplinary methods from computer science and engineering as well as cognitive science, social science and human studies evaluation, to endow robots with the ability to assist in mitigating critical societal problems that require sustained personalized support to supplement the efforts of parents, caregivers, clinicians and educators. Matarić is a Fellow of the American Association for the Advancement of Science (AAAS), Fellow of the IEEE and AAAI, recipient of the Presidential Awards for Excellence in Science, Mathematics & Engineering Mentoring (PAESMEM), the Anita Borg Institute Women of Vision Award for Innovation, Okawa Foundation Award, NSF Career Award, the MIT TR35 Innovation Award, the IEEE Robotics and Automation Society Early Career Award and has received many other awards and honors. She was featured in the science documentary movie “Me & Isaac Newton”, in The New Yorker (“Robots that Care” by Jerome Groopman, 2009), Popular Science (“The New Face of Autism Therapy”, 2010), the IEEE Spectrum (“Caregiver Robots”, 2010), and is one of the LA Times Magazine 2010 Visionaries. Matarić is the author of a popular introductory robotics textbook, “The Robotics Primer” (MIT Press 2007), an associate editor of three major journals and has published extensively.
Published: 23 March 2020
Abstract:
<p><strong>Key Words</strong>: Uncertainty Quantification, Deep Learning, Space-Time POD, Flood Modeling</p><p><br>While impressive results have been achieved in the well-known fields where Deep Learning allowed for breakthroughs such as computer vision, language modeling, or content generation [1], its impact on different, older fields is still vastly unexplored. In computational fluid dynamics and especially in Flood Modeling, many phenomena are very high-dimensional, and predictions require the use of finite element or volume methods, which can be, while very robust and tested, computational-heavy and may not prove useful in the context of real-time predictions. This led to various attempts at developing Reduced-Order Modeling techniques, both intrusive and non-intrusive. One late relevant addition was a combination of Proper Orthogonal Decomposition with Deep Neural Networks (POD-NN) [2]. Yet, to our knowledge, in this example and more generally in the field, little work has been conducted on quantifying uncertainties through the surrogate model.<br>In this work, we aim at comparing different novel methods addressing uncertainty quantification in reduced-order models, pushing forward the POD-NN concept with ensembles, latent-variable models, as well as encoder-decoder models. These are tested on benchmark problems, and then applied to a real-life application: flooding predictions in the Mille-Iles river in Laval, QC, Canada.<br>For the flood modeling application, our setup involves a set of input parameters resulting from onsite measures. High-fidelity solutions are then generated using our own finite-volume code CuteFlow, which is solving the highly nonlinear Shallow Water Equations. The goal is then to build a non-intrusive surrogate model, that&#8217;s able to <em>know what it know</em>s, and more importantly, <em>know when it doesn&#8217;t</em>, which is still an open research area as far as neural networks are concerned [3].</p><p><br><strong>REFERENCES</strong><br>[1] C. Szegedy, S. Ioffe, V. Vanhoucke, and A. A. Alemi, &#8220;Inception-v4, inception-resnet and the impact of residual connections on learning&#8221;, in Thirty-First AAAI Conference on Artificial Intelligence, 2017.<br>[2] Q. Wang, J. S. Hesthaven, and D. Ray, &#8220;Non-intrusive reduced order modeling of unsteady flows using artificial neural networks with application to a combustion problem&#8221;, Journal of Computational Physics, vol. 384, pp. 289&#8211;307, May 2019.<br>[3] B. Lakshminarayanan, A. Pritzel, and C. Blundell, &#8220;Simple and scalable predictive uncertainty estimation using deep ensembles&#8221;, in Advances in Neural Information Processing Systems, 2017, pp. 6402&#8211;6413.</p>
, Giovanni Luca Masala, Bruno Golosio, Angelo Cangelosi
Published: 16 June 2020
Frontiers in Robotics and AI, Volume 7; https://doi.org/10.3389/frobt.2020.00069

Abstract:
Editorial on the Research TopicLanguage Representation and Learning in Cognitive and Artificial Intelligence Systems In recent years, the rise of deep learning has transformed the field of Natural Language Processing (NLP), thus, producing models based on neural networks with impressive achievements in various tasks, such as language modeling (Devlin et al., 2019), syntactic parsing (Pota et al., 2019), machine translation (Artetxe et al., 2017), sentiment analysis (Fu et al., 2019), and question answering (Zhang et al., 2019). This progress has been accompanied by a myriad of new end-to-end neural network architectures able to map input text to some output prediction. On the other hand, architectures inspired by human cognition have recently appeared (Dominey, 2013; Hinaut and Dominey, 2013; Golosio et al., 2015), this is aimed at modeling language comprehension and learning by means of neural models built according to current knowledge on how verbal information is stored and processed in the human brain. Despite the success of deep learning in different NLP tasks and the interesting attempts of cognitive systems, natural language understanding still remains an open challenge for machines. The goal of this Research Topic is to describe novel and very interesting theoretical studies, models, and case studies in the areas of NLP as well as Cognitive and Artificial Intelligence (AI) systems, based on knowledge and expertise coming from heterogeneous but complementary disciplines (machine/deep learning, robotics, neuroscience, psychology). Stille et al. propose a large-scale neural model, including cognitive and lexical levels of the human neural system, with the aim of simulating the human behavior occurring in medical screenings. The large-scale neural model is biologically inspired and built by exploiting the Neural Engineering Framework and the Semantic Pointer Architecture. The authors simulate parts of both the screenings, using either the normal neural model or the neural model including neural deficits. The simulated screenings are focused on the detection of developmental problems in lexical storage and retrieval, as well as of mild cognitive impairment and early dementia. Jacobs proposes a heuristic tool called SentiArt for realizing different sentiment analyses for text segments and figures. The tool uses vector space models together with theory-guided and empirically validated label lists to compute the valence of each word in a text by locating its position in a 2d emotion potential space spanned by the >2 million words of the vector space model. By means of two computational poetics studies, the author experimentally shows the ability of SentiArt to determine the emotion of text passages and to compute emotional figure profiles and personality figure profiles for main characters from the book series (stories, novels, plays, or ballads). Ferrone and Zanzotto describe a survey aimed to deeply investigate the link between symbolic and distributed/distributional representations of Natural Language. In particular, the survey describes the general concept of representation, the notion of concatenative composition and the difference between local and distributed representations. Furthermore, it deeply addresses the general issue of compositionality, analyzing three different approaches: compositional distributional semantics, holographic reduced representations and recurrent neural networks. Nakashima et al. presents a new unsupervised machine learning method for phoneme and word discovery from multiple speakers. Human infants can acquire knowledge of phonemes and words from interactions with their mother as well as with others surrounding them. Authors propose a phoneme and word discovery method that simultaneously uses non-parametric Bayesian double articulation analyzer and deep sparse autoencoder with parametric bias in a hidden layer. Their system reduces the negative effect of speaker-dependent acoustic features in an unsupervised manner by using a speaker index required to be obtained through another speaker recognition method. This can be regarded as a more natural computational model of phoneme and word discovery by humans, because it does not use transcription. Wallbridge et al. proposes a dynamic method of communication between robots and humans in order to generate Spatial Referring Expressions describing a location. The focus of most algorithms for generation is to create a non-ambiguous description, but this is not how people naturally communicate. The authors call dynamic description how humans tend to give an underspecified description and then rely on a strategy of repair to reduce the number of possible locations or objects until the correct one is identified. The authors present a method for generating these dynamic descriptions for Human Robot Interaction, using machine learning to generate repair statements in a two-dimensional environment (game-like scenario). Miyazawa et al. presents a unified framework, integrating a cognitive architecture in a real robot for the simultaneously comprehension of concepts, actions, and language. Their integration is based on various cognitive modules and leveraging mainly multimodal categorization by using multilayered multimodal latent Dirichlet allocation (mMLDA). The integration of reinforcement learning and mMLDA enables actions based on understanding. Furthermore, the mMLDA, in conjunction with grammar learning and based on the Bayesian hidden Markov model, allows the robot to verbalize its own actions and understand user utterances. Decision making and language understanding by using abstracted concepts are verified using a real robot. Despite relevant progress having been made in the field of AI applied to NLP in the last decade, the goal of creating truly human-like intelligent systems still seems very distant. The difficulties encountered in the development of the most recent systems clearly show that the problem of human-machine interaction through natural language can no longer be addressed as a simple input-output problem. To make a qualitative leap, AI systems should become more complete multimodal systems which are able to integrate skills in areas of AI that are currently treated separately and should be capable of developing an internal representation of the external world through the combination of other information besides the verbal one. Such combination can be achieved through the integration of AI systems in robots. Embodied architectures in robots should be able to learn in a similar way to humans through interaction with humans themselves and be capable of proactively adapting operating in the environment to find the information necessary to learn and to interact more profitably with humans. From this perspective, perhaps the approaches inspired by neuroscience and cognitive models can still provide new important ideas to this field. All authors contributed equally to manuscript writing, read, and approved the final version. The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Artetxe, M., Labaka, G., Agirre, E., and Cho, K. (2017). Unsupervised neural machine translation. arXiv preprint arXiv:1710.11041. doi: 10.18653/v1/D18-1399 CrossRef Full Text | Google Scholar Devlin, J., Chang, M. W., Lee, K., and Toutanova, K. (2019). “BERT: Pre-training of deep bidirectional transformers for language understanding.” in Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, 1:4171–86. doi: 10.18653/v1/N19-1423 CrossRef Full Text | Google Scholar Dominey, P. F. (2013). Recurrent temporal networks and language acquisition—from corticostriatal neurophysiology to reservoir computing. Front. Psychol. 4:500. doi: 10.3389/fpsyg.2013.00500 PubMed Abstract | CrossRef Full Text | Google Scholar Fu, X., Wei, Y., Xu, F., Wang, T., Lu, Y., Li, J., et al. (2019). Semi-supervised aspect-level sentiment classification model based on variational autoencoder. Knowledge Based Syst. 171, 81–92. doi: 10.1016/j.knosys.2019.02.008 CrossRef Full Text | Google Scholar Golosio, B., Cangelosi, A., Gamotina, O., and Masala, G. L. (2015). A cognitive neural architecture able to learn and communicate through natural language. PLoS ONE 10:e0140866. doi: 10.1371/journal.pone.0140866 PubMed Abstract | CrossRef Full Text | Google Scholar Hinaut, X., and Dominey, P. F. (2013). Real-time parallel processing of grammatical structure in the fronto-striatal system: a recurrent network simulation study using reservoir computing. PLoS ONE 8:e52946. doi: 10.1371/journal.pone.0052946 PubMed Abstract | CrossRef Full Text | Google Scholar Pota, M., Marulli, F., Esposito, M., De Pietro, G., and Fujita, H. (2019). Multilingual POS tagging by a composite deep architecture based on character-level features and on-the-fly enriched Word Embeddings. Knowledge Based Syst. 164, 309–323. doi: 10.1016/j.knosys.2018.11.003 CrossRef Full Text | Google Scholar Zhang, Z., Wu, Y., Zhou, J., Duan, S., and Zhao, H. (2019). SG-Net: Syntax-guided machine reading comprehension. The Thirty-Fourth AAAI Conference on Artificial Intelligence, AAAI, 2020, The Thirty-Second Innovative Applications of Artificial Intelligence Conference, IAAI 2020, The Tenth AAAI Symposium on Educational Advances in Artificial Intelligence, EAAI 2020, (New York, NY), 9636–43. Google Scholar Keywords: Natural Language Processing (NLP), artificial intelligence, cognitive systems, robotics, deep learning, machine learning, language representation and language processing Citation: Esposito M, Masala GL, Golosio B and Cangelosi A (2020) Editorial: Language Representation and Learning in Cognitive and Artificial Intelligence Systems. Front. Robot. AI 7:69. doi: 10.3389/frobt.2020.00069 Received: 13 February 2020; Accepted: 27 April 2020; Published: 16 June 2020. Edited and reviewed by: Mikhail Prokopenko, University of Sydney, Australia Copyright © 2020 Esposito, Masala, Golosio and Cangelosi. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. *Correspondence: Massimo Esposito, [email protected]
Andres Garcia-Saavedra, Xi Li, Alexandros Kostopoulos, George Iosifidis, Danny De Vleeschauwer, Chia-Yu Chang, Lidia Fuentes, Daniel-Jesus Munoz, Brendan McAuliffe, Paul Sutton, et al.
Published: 29 March 2021
by 10.5281
The publisher has not yet granted permission to display this abstract.
Andres Garcia-Saavedra, Xi Li, Alexandros Kostopoulos, George Iosifidis, Danny De Vleeschauwer, Chia-Yu Chang, Lidia Fuentes, Daniel-Jesus Munoz, Brendan McAuliffe, Paul Sutton, et al.
Published: 29 March 2021
by 10.5281
The publisher has not yet granted permission to display this abstract.
Blanca Cabrera Gil, Laura Gui Levy, Mario Kreutzfeldt, Joanna Kowal, Giusy Procopio, Deniz Eroglu, Doron Merkler, Andrew Janowczyk, Diego Dupouy, Aitana Neves, et al.
Published: 7 November 2022
by BMJ
Abstract:
Background Colorectal carcinoma (CRC) represents a major worldwide health burden and shows an increasing incidence particularly in younger patients. The majority of metastatic CRC (microsatellite stable, MSS) do not respond to current immunotherapies in contrast to the small subset of microsatellite unstable (MSI) patients. However, responses to immunotherapy were recently observed in subsets of primary MSS tumors indicating their potential vulnerability to such therapeutic approaches. Here, we use multiplex imaging coupled with powerful image analysis tools to highlight important phenotypic differences between MSI and MSS. Methods 100 human CRC sections (30MSI, 70 MSS) were examined on the COMET™ platform (Lunaphore) with a multiplex sequential immunofluorescence (seqIF) panel of 12 biomarkers (CD3, CD31, CD4, CD8, FoxP3, TCF-1, TOX, EOMES, CK, CD45RO, S100, D2-40). Images were preprocessed by background subtraction and local contrast enhancement.1 Cell segmentation was done with the Stardist algorithm.2Cell phenotypes (CT) were defined by threshold-based classifiers. The epithelium was detected based on CK expression. The density of CT was assessed per area of interest. Cell neighborhoods (CN) were defined as the composition of the 10 closest CTs within 300um. CNs were clustered into CN-classes using DB-SCAN.3Results Cell segmentation and phenotyping algorithms resulted in precise cell detection (figure 1). We compared CT infiltration patterns in MSI vs MSS in the tumor-epithelium and epithelium-stroma interfaces. While we observed no significant differences in CT densities at the interface, CD3 cell density was significantly higher in MSI vs MSS patients in the tumor-epithelium (figure 2), with MSS more heterogeneously distributed with CD3-high and CD3-low outliers. Differences in cell interactions CRC patients were examined using CN analysis. We observed a CN composed of CD3, CD8 cytotoxic T cells, CD3+CD8+ and tumor cells to be differentially enriched between MSI and MSS patients (figure 3). Conclusions A more precise characterization of the immune response to cancers is essential to leverage the advantages of immune modulations in the treatment of cancers. We confirm here important differences between MSI and MSS CRCs in part systematic due to the higher antigen load in MSI CRCs but also highlight heterogeneity amongst the MSS tumors. Characterization of the phenotype of the T-lymphocyte population and its localization with respect to elements such as epithelial cells and tertiary lymphoid structures may help to define both the prognosis of tumors and the possibility of a response to immune checkpoint therapy. References SM Pizer, et al. Computer Vision, Graphics, and Image Processing 39, 1987. U Schmidt, et al. MICCAI, 2018. Martin Ester, Hans-Peter Kriegel, Jörg Sander, and Xiaowei Xu. 1996. A density-based algorithm for discovering clusters in large spatial databases with noise. In Proceedings of the Second International Conference on Knowledge Discovery and Data Mining (KDD’96). AAAI Press, 226–231. Massey FJ 1951. The Kolmogorov-Smirnov Test for Goodness of Fit. Journal of the American Statistical Association. 1951;46(253):68–78. Cohen J 1988. Statistical power analysis for the behavioral sciences (2nd ed.). Hillside, NJ: Lawrence Erlbaum Associates. Ethics Approval The study has been approved by HUG. Example results from cell segmentation and phenotyping. Left: Original image patch. Blue: DAPI, green: CD4, red: CD8. Right: Patch with nuclei segmentation (all contours), CD4+ cells (cyan) and CD8+ cells (yellow) CT analysis of the tumor epithelium region. Each boxplot represents the cell density distribution of a CT in MSS and MSI patients. CD3 cell density is observed to be significantly higher in MSI vs MSS patients, with MSS more heterogeneously distributed with CD3-high and CD3-low outliers Cell neighborhood-class consisting of CD3, CD8, and CD3_CD8 cell types is significantly more frequent in MSI vs MSS patients. a) CN frequency in MSI vs MSS patients (KS-test [4] p-value and Cohen d [5] effect size). b) Proportion of CTs in this CN-class
, Shaoyuan Li
Published: 2 November 2020
Control Theory and Technology, Volume 18, pp 459-461; https://doi.org/10.1007/s11768-020-00008-w

The publisher has not yet granted permission to display this abstract.
, Umut Oztok
Published: 9 November 2012
Journal: EMBnet.journal
EMBnet.journal, Volume 18; https://doi.org/10.14806/ej.18.b.551

Abstract:
Motivation and Objectives Storing biomedical data in various structured forms, like biomedical databases or ontologies, and at different locations have brought about many challenges for answering complex queries about the knowledge represented in these resources. For instance, here are two queries about some genes, drugs and diseases: “What are the drugs that treat the disease Depression and that do not target the gene ACYP1?”, “What are the 3 most similar drugs that target the gene DLG4?” One of the challenges of answering such complex queries is to represent the queries in a natural language and present the answers in an understandable form. Another challenge is to efficiently find answers to complex queries that require appropriate integration of relevant knowledge stored in different places and in various forms, and/or that require auxiliary definitions, such as, chains of drug-drug interactions, cliques of genes based on gene-gene relations, similarity/diversity of genes/drugs. Furthermore, once an answer is found for a complex query, the experts may need further explanations about the answer. We have developed novel computational methods and built a software system, called BioQuery-ASP, to handle all these challenges Methods We have addressed the challenges described above using a declarative programming paradigm, called Answer Set Programming (ASP) (Lifschitz, 2008; Brewka et al., 2011). ASP provides an expressive high-level knowledge representation formalism that allows recursive definitions, aggregates, default negation, etc. and efficient automated reasoners, such as Clasp (Gebser et al., 2007), which has recently won first places at ASP and SAT (Boolean Satisfiability) competitions in automated reasoning. Due to these attractive features, ASP has been used in various applications, such as phylogeny reconstruction (Brooks et al., 2006), systems biology (Gebser et al., 2011), service robotics (Aker et al., 2012), decision support systems (Nogueira et al., 2001), automatic music construction (Boenn et al., 2009), workforce management (Ricca et al., 2012). To address the first challenge (i.e., representing queries in natural language), we have developed a controlled natural language (called BioQuery-CNL) for biomedical queries about drug discovery (Erdem and Yeniterzi, 2009; Oztok 2012). For instance, the queries above are in BioQuery-CNL. Then we have built an intelligent user interface that allows users to enter biomedical queries in BioQuery-CNL and that presents the answers with links to related webpages (Erdem et al., 2011b). Queries in BioQuery-CNL are translated into a set of ASP rules by a novel algorithm. For instance, the first query above is translated into the following ASP rules: what _ drug(DRG) <- drug _ name(DRG), drug _ treats _ disease(DRG,"Depression"), not drug _ targets _ gene(DRG,"ACYP1") which describe the drugs DRG that treat the disease Depression and that do not target the gene ACYP1. To address the second challenge (i.e., efficiently answering complex queries), first we have developed a rule layer over biomedical ontologies and databases that not only integrates the concepts in these knowledge resources but also provides definitions of auxiliary concepts (Bodenreider et al., 2008). For instance, the predicate drug_treats_disease is defined in the rule layer as follows: drug _ treats _ disease(DRG,DIS) <- drug _treats _ disease _ pkb(DRG,DIS) drug _ treats _ disease(DRG,DIS) <- drug _treats _ disease _ ctd(DRG,DIS) integrating the knowledge extracted from the knowledge bases PharmGKB (McDonagh et al., 2011) and CTD (Davis et al., 2011), about "which drug treats which disease." The auxiliary concept of "chains of gene-gene relations" is defined recursively in the rule layer as well: gene _ reachable _ from(X,1) <- gene _gene(X,Y), start _ gene(Y) gene _ reachable _ from(X,N+1) <- gene _gene(X,Z), gene _ reachable _ from(Z,N), N < L, max _chain _ length(L) to be able to answer queries like "What are the genes related to the gene ADRB1 via a gene-gene relation chain of length at most 3?" Then, for an efficient query answering, we have introduced an algorithm to identify the relevant parts of the rule layer and the knowledge resources with respect to the given query, and used automated reasoners of ASP to answer queries considering these relevant parts (Erdem et al., 2011a). Essentially, our algorithm identifies the relevant predicates that the query-predicates depend on (using a "dependency graph"), and considers the rules that contain these relevant predicates. For some queries, the relevant knowledge consists of about 500 thousand rules whereas the total size of all the knowledge resources (with the rule layer) is over 21 million rules; considering the relevant rules only decreases the computation time of answering a query by almost a factor of 100. To address the third challenge (i.e., generating explanations), we have developed an intelligent algorithm to generate an explanation (i.e., a tree of "applicable" ASP rules) for a given answer, with respect to the query and the relevant parts of the rule layer and the knowledge resources. We have also developed algorithms to generate shortest/different explanations for a biomedical query taking into account the provenance information as well (Oztok 2012). For instance, an answer to the query "What are the genes that are targeted by the drug Epinephrine and that interact with the gene DLG4?" is ADRB1; and a shortest explanation that justifies this answer is as follows: "The drug Epinephrine targets the gene ADRB1 according to CTD and the gene DLG4 interacts with the gene ADRB1 according to BioGrid." Based on these methods, we have developed a software system, BioQuery-ASP, that guides the user to represent a complex query in a natural language, finds answers to the query (if an answer exists), returns links to related web pages for further information, and generates explanations (if the user asks for one). A demo of BioQuery-ASP is available at BioQuery-ASP Website: http://krr.sabanciuniv.edu/projects/BioQuery-ASP/ (Last accessed on September 25, 2012)). Results and Discussion We have shown the applicability of BioQuery-ASP to answer complex queries that are specified by experts, over large biomedical knowledge resources about genes, drugs and diseases, such as PharmGKB, DrugBank (Knox et al., 2011), BioGrid (Stark et al., 2006), CTD, Sider (Kuhn et al., 2010), etc., using efficient solvers of ASP. BioQuery-ASP could find answers to most of the complex queries in 3-10 CPU seconds, over 10 million facts extracted from these knowledge resources and over 10 million rules integrating them (using a computer with two 1.60GHz Intel Xeon E5310 Quad-core Processors and 16GB RAM). No existing biomedical query answering systems (e.g., web services built over the available knowledge resources, which answer queries by means of keyword search) can directly answer such queries, or can generate explanations for answers. In that sense, BioQuery-ASP is a novel biomedical query answering system that can be useful for experts in automating deep reasoning about knowledge about genes, drugs and diseases available via various biomedical databases and ontologies. Acknowledgements This work has been partly supported by TUBITAK Grant 108E229. References Aker E, Patoglu V, et al. (2012) Answer Set Programming for Reasoning with Semantic Knowledge in Collaborative Housekeeping Robotics. In Proc. of the 10th IFAC Symposium on Robot Control. Bodenreider O, Coban Z, et al. (2008) A preliminary report on answering complex queries related to drug discovery using answer set programming. In Proc. of the 3rd International Workshop on Applications of Logic Programming to the Semantic Web and Web Services. Boenn G, Brain M, et al. (2009) Anton: Composing logic and logic composing. In Proc. of the 10th International Conference on Logic Programming and Nonmonotonic Reasoning, pages 542-547. Brewka G, Eiter T, et al. (2011) Answer set programming at a glance. Communications of ACM 54(12):92-103. Brooks DR, Erdem E, et al. (2006) Inferring Phylogenetic Trees Using Answer Set Programming. Journal of Automated Reasoning 39(4): 471-511. Davis AP, King BL, et al. (2011) The Comparative Toxicogenomics Database: update 2011. Nucleic Acids Research 39(Database issue):D1067-72. Erdem E, Erdem Y, et al. (2011a) Finding answers and generating explanations for complex biomedical queries. In Proc. of the 25th Conf. on Artificial Intelligence (AAAI), pages 785-790. Erdem E, Erdogan H, et al. (2011b) BioQuery-ASP: Querying biomedical ontologies using answer set programming. In Proc. of [email protected] Challenge. Erdem E and Yeniterzi R (2009) Transforming controlled natural language biomedical queries into answer set programs. In Proc. of BioNLP Workshop, pages 117-124. Gebser M, Kaufmann B, et al. (2007) clasp: A Conflict-Driven Answer Set Solver. In Proc. of the 9th Int'l Conference on Logic Programming and Nonmonotonic Reasoning, pages 260-265. Gebser M, Schaub T, et al. (2011) Detecting inconsistencies in large biological networks with answer set programming. Theory and Practice of Logic Programming, 11(2):1-38. Knox C, Law V, et al. (2011) DrugBank 3.0: a comprehensive resource for 'omics' research on drugs. Nucleic Acids Research 39(Database issue):D1035-41. Kuhn M, Campillos M, et al. (2010) A side effect resource to capture phenotypic effects of drugs. Molecular Systems Biology 6:343. Lifschitz V (2008) What Is Answer Set Programming? In Proc. of the 23rd Conference on Artificial Intelligence (AAAI), pages 1594-1597. McDonagh EM, Whirl-Carrillo M, et al. (2011) From pharmacogenomic knowledge acquisition to clinical applications: the PharmGKB as a clinical pharmacogenomic biomarker resource. Biomarkers in Medicine 5(6):795-806. Nogueira M, Balduccini M, et al. (2001) An A-Prolog decision support system for the space shuttle. In Proc. of the 3rd Int'l Symposium on Practical Aspects of Declarative Languages, pages 169-183. Oztok U (2012) Generating Explanations for Complex Biomedical Queries. M.S. Thesis, Sabanci University. Stark C, Breitkreutz BJ, et al. (2006) BioGRID: a general repository for interaction datasets. Nucleic Acids Research 34(Database issue):D535-9. Ricca F, Grasso G,et al. (2012) Team-building with answer set programming in the Gioia-Tauro seaport. Theory and Practice of Logic Programming 12(3):361-381.
, Huajin Tang, Kaushik Roy
Published: 17 March 2021
Frontiers in computational neuroscience, Volume 15; https://doi.org/10.3389/fncom.2021.665662

Abstract:
Editorial on the Research TopicUnderstanding and Bridging the Gap between Neuromorphic Computing and Machine Learning On the road toward artificial general intelligence (AGI), two solution paths have been explored: neuroscience-driven neuromorphic computing such as spiking neural networks (SNNs) and computer-science-driven machine learning such as artificial neural networks (ANNs). Owing to availability of data, high-performance processors, effective learning algorithms, and easy-to-use programming tools, ANNs have achieved tremendous breakthroughs in many intelligent applications. Recently, SNNs also attracted a lot of attention due to its biological plausibility and the possibility of achieving energy-efficiency (Roy et al., 2019). However, they suffer from ongoing debates and skepticisms due to worse accuracy compared to “standard” ANNs. The performance gap comes from a variety of factors, including learning techniques, benchmarks, programming tools and execution hardware, all of which in SNNs are not as developed as those in the ANN domain. To this end, we propose a Research Topic, named “Understanding and Bridging the Gap between Neuromorphic Computing and Machine Learning,” in Frontiers in Neuroscience and Frontiers in Computational Neuroscience to collect recent researches on neuromorphic computing and machine learning to help understand and bridge the aforementioned gap. We received 18 submissions in total and accepted 14 of them in the end. The scope of these accepted papers covers learning algorithms, applications, and efficient hardware. How to train SNN models is the key to improve its functionality, thus bridging the gap between ANN models. Unlike the ANN domain that has grown rapidly via sophisticated backpropagation-based learning algorithms, the SNN domain is still short of effective learning algorithms due to the complicated spatiotemporal dynamics and non-differentiable spike activities. Currently, there are overall two categories of learning algorithms for SNNs: unsupervised synaptic plasticity with biological plausibility [e.g., spike timing dependent plasticity, STDP (Diehl and Cook, 2015)] and supervised backpropagation with gradient descent [e.g., indirect learning by acquiring gradients from the ANN counterpart (Diehl et al., 2015; Sengupta et al., 2019), direct learning by acquiring gradients from the SNN itself (Lee et al., 2016; Wu et al., 2018; Gu et al., 2019; Zheng et al., 2021), or the combination of both (Rathi et al., 2020)]. The latter can usually achieve higher accuracy and has advanced the model scale to handle ImageNet-level tasks. In the future, we look forward to seeing more studies on SNN learning to close the gap. Next, we briefly summarize the recent progress of neural network (especially SNN) learning presented in our accepted papers. Inspired by the curiosity-based learning mechanism of the brain, Shi et al. propose curiosity-based SNN (CBSNN) models. In the first training epoch, the novelty estimations of all samples are obtained through bio-plausible synaptic plasticity; next, the samples whose novelty estimations exceed the threshold are repeatedly learned and the novelty estimations are updated in the next five epochs; then, all samples are learned with one more epoch. The last two steps are periodically taken until convergence. CBSNNs show better accuracy and higher efficiency in processing several small-scale datasets than conventional voltage-driven plasticity-centric SNNs. Daram et al. propose ModNet, an efficient dynamic learning system inspired from the neuromodulatory mechanism in the insect brain. An inbuilt modulatory unit regulates learning based on the context and internal state of the system. The network with modulatory trace achieves 98.8% ± 1.16 on average over the omniglot dataset for five-way one-shot image classification task while using 20x fewer trainable parameters compared to state-of-the-art models. Kaiser, Mostafa et al. introduce deep continuous local learning (DECOLLE), an SNN model equipped with local error functions for online learning. The synaptic plasticity rules are derived from user-defined cost functions and neural dynamics by leveraging existing autodifferentiation methods of machine learning frameworks. The model demonstrates state-of-the-art performance on N-MNIST and DvsGesture datasets. Fang et al. propose a bio-plausible noise structure to optimize the performance of SNNs trained by gradient descent. Through deducing the strict saddle condition for synaptic plasticity, they demonstrate that the noise helps the optimization escape from saddle points on high dimensional domains. The accuracy improvement can reach at least 10% on MNIST and CIFAR10 datasets. Panda et al. modify the SNN configuration with backward residual connections, stochastic softmax, and hybrid artificial-and-spiking neuronal activations. In this way, the previous learning methods are improved with comparable accuracy but large efficiency gains over the ANN counterparts. Unlike the artificial neuron in ANNs, each spiking neuron in SNNs has intrinsic temporal dynamics, which is appropriate for processing sequence information. In this Research Topic, we accepted two papers that discuss SNN applications. Wu et al. explore the first work that uses SNNs for large-vocabulary continuous automatic speech recognition (LVCSR) tasks. Their SNNs demonstrate competitive accuracies on par with their ANN counterparts while consuming only 10 algorithmic timesteps and 0.68× total synaptic operations. They integrate the models into the PyTorch-Kaldi Speech Recognition Toolkit for rapid development. Kugele et al. apply SNNs for processing spatiotemporal event streams (e.g., N-MNIST, CIFAR10-DVS, N-CARS, and DvsGesture datasets). They improve the ANN-to-SNN conversion learning method by introducing connection delays during the pre-training of ANNs to match the propagation delays in converted SNNs. In this way, the resulting SNNs can handle the above tasks accurately and efficiently. In addition, besides energy-efficiency (Merolla et al., 2014), recent studies further find that the event-driven computing paradigm of SNNs endows them high robustness (He et al., 2020; Liang et al., 2020) and superior capability in learning sparse features (He et al., 2020). We believe it is very important to mine the true advantages of SNNs to determine their true value in practical applications. Performing neural networks on general-purpose processors is inefficient, which stimulates the development of various domain-specific hardware platforms, including those for ANNs [e.g., DaDianNao (Chen et al., 2014), TPU (Jouppi et al., 2017), Eyeriss (Chen et al., 2017), Thinker (Yin et al., 2017), etc.), for SNNs (e.g., SpiNNaker (Furber et al., 2014), TrueNorth (Merolla et al., 2014), Loihi (Davies et al., 2018), DYNAPs (Moradi et al., 2017)], and for cross-paradigm modeling [e.g., Tianjic (Pei et al., 2019; Deng et al., 2020)]. In this Research Topic, we accepted seven papers for neural network hardware: three for ANNs, two for SNNs, and two for cross-paradigm. Sim and Lee propose SC-CNN, the bitstream-based convolutional neural network (CNN) inspired by stochastic computing (SC) that uses bitstreams to represent numbers, to improve machine learning hardware. Benefitting from the CNN substrate, SC-CNN can achieve high accuracy; further benefitting from SC, SC-CNN is highly efficient, scalable, and fault-tolerant. Different from the common digital machine learning accelerators, Kaiser, Faria et al. present a clockless autonomous probabilistic circuit, wherein synaptic weights are read out in the form of analog voltages, for fast and efficient learning with no use of digital computing. They demonstrate a circuit built with existing technology to emulate the Boltzmann machine learning algorithm. Muller et al. introduce bias matching, a top-down neural network design approach, to match the inductive biases required in a machine learning system to the hardware constraints of its implementation. To alleviate the high cost training of SNNs using backpropagation, Lee et al. propose a spike-train level direct feedback alignment (ST-DFA) algorithm. Compared to the state-of-the-art backpropagation learning algorithm, they demonstrate excellent performance vs. overhead tradeoffs on FPGA for speech and image classification applications. Dutta et al. propose an all ferroelectric field-effect transistors (FeFET)-based SNN hardware that allows low-power spike-based information processing and co-localized memory and computing. They implement a surrogate gradient supervised learning algorithm on their efficient SNN platform, which further accounts for the impacts of device variation and limited bit precision of on-chip synaptic weights on the classification accuracy. Parsa et al. build a hierarchical pseudo agent-based multi-objective Bayesian hyperparameter optimization framework for both software and hardware. They can not only maximize the performance of the network, but also minimize the energy and area overheads of the corresponding neuromorphic hardware. They validate the proposed framework using both ANN and SNN models, which involves both deep learning accelerators [e.g., PUMA (Ankit et al., 2019)] and neuromorphic hardware [e.g., DANNA2 (Mitchell et al., 2018) and mrDANNA (Chakma et al., 2017)]. Instead of implementing ANNs and SNNs separately, integration of them has become a promising direction to achieve further breakthroughs toward AGI via complementary advantages (Pei et al., 2019). Therefore, the efficient hardware that can support individual modeling of ANNs and SNNs as well as their hybrid modeling is very important. This has been achieved by the cross-paradigm Tianjic platform (Deng et al., 2020), based on which Wang et al. further present an end-to-end mapping framework for implementing various hybrid neural networks. By constructing hardware configuration schemes for four typical signal conversions and establishing a global timing adjustment mechanism among different heterogeneous modules, they implement hybrid models with low execution latency and low power consumption. Machine learning and neuromorphic computing are two modeling paradigms on the road toward AGI. ANNs have achieved tremendous breakthroughs in many intelligent applications benefitting from big data, high-performance processors, effective learning algorithms, and easy-to-use programming tools; in contrast, SNNs are still in its infant stage and there is a dire need for more neuromorphic benchmarks. Through cross-discipline research, we expect to understand and bridge the gap between neuromorphic computing and machine learning. This Research Topic is just a small step in this direction, and we look forward to more innovations that can achieve brain-like intelligence. All authors listed have made a substantial, direct and intellectual contribution to the work, and approved it for publication. This work was supported in part by the key scientific technological innovation research project by Ministry of Education of China, Zhejiang Lab (Grant No. 2019KC0AD02), Center for Brain Inspired Computing (C-BRIC), a Semiconductor Research Corporation (SRC) program sponsored by DARPA, National Science Foundation, Sandia National Laboratory, ONR sponsored Multi-University Research Initiative (MURI), and Vannevar Bush Faculty Fellowship program. The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Ankit, A., Hajj, I. E., Chalamalasetti, S. R., Ndu, G., Foltin, M., Williams, R. S., et al. (2019). “PUMA: A programmable ultra-efficient memristor-based accelerator for machine learning inference,” in Proceedings of the Twenty-Fourth International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS) (Providence, RI: ACM), 715–731. doi: 10.1145/3297858.3304049 CrossRef Full Text | Google Scholar Chakma, G., Adnan, M. M., Wyer, A. R., Weiss, R., Schuman, C. D., and Rose, G. S. (2017). Memristive mixed-signal neuromorphic systems: Energy-efficient learning at the circuit-level. IEEE J. Emerg. Selected Topics Circuits Syst. 8, 125–136. doi: 10.1109/JETCAS.2017.2777181 CrossRef Full Text | Google Scholar Chen, Y.-H., Krishna, T., Emer, J. S., and Sze, V. (2017). Eyeriss: An energy-efficient reconfigurable accelerator for deep convolutional neural networks. IEEE J. Solid-State Circuits 52, 127–138. doi: 10.1109/JSSC.2016.2616357 CrossRef Full Text | Google Scholar Chen, Y., Luo, T., Liu, S., Zhang, S., He, L., Wang, J., et al. (2014). “Dadiannao: a machine-learning supercomputer,” in 47th Annual IEEE/ACM International Symposium on Microarchitecture (MICRO) (Cambridge: IEEE/ACM), 609–622. doi: 10.1109/MICRO.2014.58 CrossRef Full Text | Google Scholar Davies, M., Srinivasa, N., Lin, T.-H., Chinya, G., Cao, Y., Choday, S. H., et al. (2018). Loihi: a neuromorphic manycore processor with on-chip learning. IEEE Micro 38, 82–99. doi: 10.1109/MM.2018.112130359 CrossRef Full Text | Google Scholar Deng, L., Wang, G., Li, G., Li, S., Liang, L., Zhu, M., et al. (2020). Tianjic: A unified and scalable chip bridging spike-based and continuous neural computation. IEEE J. Solid-State Circuits 55, 2228–2246. doi: 10.1109/JSSC.2020.2970709 CrossRef Full Text | Google Scholar Diehl, P. U., and Cook, M. (2015). Unsupervised learning of digit recognition using spike-timing-dependent plasticity. Front. Comput. Neurosci. 9:99. doi: 10.3389/fncom.2015.00099 PubMed Abstract | CrossRef Full Text | Google Scholar Diehl, P. U., Neil, D., Binas, J., Cook, M., Liu, S.-C., and Pfeiffer, M. (2015). “Fast-classifying, high-accuracy spiking deep networks through weight and threshold balancing," in International Joint Conference on Neural Networks (IJCNN) (Killarney: IEEE), 1–8. doi: 10.1109/IJCNN.2015.7280696 CrossRef Full Text | Google Scholar Furber, S. B., Galluppi, F., Temple, S., and Plana, L. A. (2014). The Spinnaker project. Proc. IEEE 102, 652–665. doi: 10.1109/JPROC.2014.2304638 CrossRef Full Text | Google Scholar Gu, P., Xiao, R., Pan, G., and Tang, H. (2019). “STCA: Spatio-temporal credit assignment with delayed feedback in deep spiking neural networks,” in International Joint Conferences on Artificial Intelligence (IJCAI), 1366–1372. doi: 10.24963/ijcai.2019/189 CrossRef Full Text | Google Scholar He, W., Wu, Y., Deng, L., Li, G., Wang, H., Tian, Y., et al. (2020). Comparing SNNs and RNNs on neuromorphic vision datasets: similarities and differences. Neural Networks 132, 108–120. doi: 10.1016/j.neunet.2020.08.001 PubMed Abstract | CrossRef Full Text | Google Scholar Jouppi, N. P., Young, C., Patil, N., Patterson, D., Agrawal, G., Bajwa, R., et al. (2017). “In-datacenter performance analysis of a tensor processing unit,” in Proceedings of the 44th Annual International Symposium on Computer Architecture (ISCA) (Toronto, ON: ACM), 1-12. doi: 10.1145/3079856.3080246 CrossRef Full Text | Google Scholar Lee, J. H., Delbruck, T., and Pfeiffer, M. (2016). Training deep spiking neural networks using backpropagation. Front. Neurosci. 10:508. doi: 10.3389/fnins.2016.00508 PubMed Abstract | CrossRef Full Text | Google Scholar Liang, L., Hu, X., Deng, L., Wu, Y., Li, G., Ding, Y., et al. (2020). Exploring adversarial attack in spiking neural networks with spike-compatible gradient. arXiv preprint arXiv:2001.01587. Google Scholar Merolla, P. A., Arthur, J. V., Alvarez-Icaza, R., Cassidy, A. S., Sawada, J., Akopyan, F., et al. (2014). A million spiking-neuron integrated circuit with a scalable communication network and interface. Science 345, 668–673. doi: 10.1126/science.1254642 PubMed Abstract | CrossRef Full Text | Google Scholar Mitchell, J. P., Dean, M. E., Bruer, G. R., Plank, J. S., and Rose, G. S. (2018). “DANNA 2: Dynamic adaptive neural network arrays,” in Proceedings of the International Conference on Neuromorphic Systems (ICONS) (Knoxville, TN: ACM), 1–6. doi: 10.1145/3229884.3229894 CrossRef Full Text | Google Scholar Moradi, S., Qiao, N., Stefanini, F., and Indiveri, G. (2017). A scalable multicore architecture with heterogeneous memory structures for dynamic neuromorphic asynchronous processors (DYNAPs). IEEE Trans. Biomed. Circuits Syst. 12, 106–122. doi: 10.1109/TBCAS.2017.2759700 PubMed Abstract | CrossRef Full Text | Google Scholar Pei, J., Deng, L., Song, S., Zhao, M., Zhang, Y., Wu, S., et al. (2019). Towards artificial general intelligence with hybrid Tianjic chip architecture. Nature 572, 106–111. doi: 10.1038/s41586-019-1424-8 PubMed Abstract | CrossRef Full Text | Google Scholar Rathi, N., Srinivasan, G., Panda, P., and Roy, K. (2020). “Enabling deep spiking neural networks with hybrid conversion and spike timing dependent backpropagation,” in International Conference on Learning Representations (ICLR). Google Scholar Roy, K., Jaiswal, A., and Panda, P. (2019). Towards spike-based machine intelligence with neuromorphic computing. Nature 575, 607–617. doi: 10.1038/s41586-019-1677-2 PubMed Abstract | CrossRef Full Text | Google Scholar Sengupta, A., Ye, Y., Wang, R., Liu, C., and Roy, K. (2019). Going deeper in spiking neural networks: VGG and residual architectures. Front. Neurosci. 13:95. doi: 10.3389/fnins.2019.00095 PubMed Abstract | CrossRef Full Text | Google Scholar Wu, Y., Deng, L., Li, G., Zhu, J., and Shi, L. (2018). Spatio-temporal backpropagation for training high-performance spiking neural networks. Front. Neurosci. 12:331. doi: 10.3389/fnins.2018.00331 PubMed Abstract | CrossRef Full Text | Google Scholar Yin, S., Ouyang, P., Tang, S., Tu, F., Li, X., Zheng, S., et al. (2017). A high energy efficient reconfigurable hybrid neural network processor for deep learning applications. IEEE J. Solid-State Circuits 53, 968–982. doi: 10.1109/JSSC.2017.2778281 CrossRef Full Text | Google Scholar Zheng, H., Wu, Y., Deng, L., Hu, Y., and Li, G. (2021). “Going deeper With directly-trained larger spiking neural networks,” in AAAI Conference on Artificial Intelligence (AAAI) (AAAI Press). Google Scholar Keywords: neuromorphic computing, spiking neural networks, machine learning, artificial neural networks, cross paradigm Citation: Deng L, Tang H and Roy K (2021) Editorial: Understanding and Bridging the Gap Between Neuromorphic Computing and Machine Learning. Front. Comput. Neurosci. 15:665662. doi: 10.3389/fncom.2021.665662 Received: 08 February 2021; Accepted: 12 February 2021; Published: 17 March 2021. Edited and reviewed by: Si Wu, Peking University, China Copyright © 2021 Deng, Tang and Roy. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. *Correspondence: Lei Deng, [email protected]
Comment
Rose Martin, Ivaylo Kusev, Alex J. Cooke, Victoria Baranova, Paul Van Schaik,
Published: 24 May 2017
Frontiers in Psychology, Volume 8; https://doi.org/10.3389/fpsyg.2017.00808

Abstract:
A commentary onThe Social Dilemma of Autonomous Vehiclesby Bonnefon, J.-F., Shariff, A., and Rahwan, I. (2016). Science 352, 1573–1576. doi: 10.1126/science.aaf2654 An autonomous vehicle (AV) car with 1 passenger (e.g., the car owner) inside is traveling within the speed limit down the road. However, unexpectedly, 10 pedestrians have appeared in its path and it is too late for the car to brake. The car must either save the passenger by driving into the 10 pedestrians and killing them, or save the 10 pedestrians by swerving into a barrier and killing the passenger. Should the AV algorithm be programmed to save the passenger, or to save the greater number of people? This question is of great importance to the AV industry, policy makers, the potential buyers, and the general public. Recent research (Bonnefon et al., 2016) has investigated how humans judge the morality of the two AV algorithms—a utilitarian (saving the greater number of lives; Bentham, 1970) and a non-utilitarian passenger-protective (saving the passenger). Using moral dilemma scenarios in which an AV is programed to be utilitarian or passenger-protective, participants were required to rate (on a 0–100 slider) “what action they thought was the most moral,” Bonnefon et al. (2016) found that participants rated the utilitarian algorithm (e.g., sacrificing 1 passenger to save 10 pedestrians) as the moral course of action. However, when the respondents were asked to rate on a scale the likelihood of buying a car with each of the algorithms (to what extent they are inclined), they indicated higher likelihood of purchasing the passenger-protective algorithm than the utilitarian one. This surprising result demonstrates what appears to be a social dilemma—an agent temptation to act in accordance with self-interest (Bonnefon et al., 2016), which often results in the worst outcome for all individuals involved, including the decision-maker (Dawes, 1980; Kollock, 1998). The authors have not explained theoretically the results from this social dilemma; yet they discounted the possibility of uncertainty (e.g., the possibility that people may not be aware or have access to the utilitarian actions and their consequences; Kusev et al., 2016; Zhao et al., 2016). Therefore, the aim of this paper is to provide an insight (theoretical and methodological) and explanation for the surprising reversals of the moral utilitarian preferences reported in Bonnefon et al. (2016). Here, we argue that methodological issues in Bonnefon's et al. (2016) research may have induced uncertainty amongst participants. Accordingly, it is plausible that the difference in response to the two questions (what action they thought was the most moral, and to what extent are they are inclined to purchase an AV with each algorithm) is caused by the restricted accessibility of moral utilitarian information. For example, respondents may not realize that a car purchaser is also inevitably a pedestrian too. We suggest that full utilitarian descriptions provide accessibility to utilitarian tasks and their consequences and can eliminate the conflicting responses to these two questions. Psychological uncertainty has been found to account for respondents' differences in utilitarian choice for morally sensitive scenarios (Kusev et al., 2016); comprehensive moral tasks and questions reduced decision uncertainty and boosted utility maximization. Kusev et al. (2016) argued that in moral decision-making tasks (the trolley and footbridge dilemma; Thomson, 1985; Greene et al., 2001) participants are given (i) a partial moral task description which outlines what will happen should they throw the switch/push the stranger, and (ii) asked a partial appropriateness of action question for only one of the two possible moral actions (yes/no answers). Hence, the respondents are left to infer what will happen should they refrain from this action and asked to judge the appropriateness of only one of the actions. Thus, the moral dilemma and question contain only “partial utilitarian descriptions,” inducing uncertainty. Accordingly, we argue that all of the scenarios presented by Bonnefon et al. (2016) contained partial utilitarian information, inducing decision uncertainty amongst respondents. For instance, in some of the scenarios each respondent is required to imagine themselves as a passenger inside the car and are therefore presented with only one side of the situation. We propose that if the respondents imagine themselves as pedestrians as well, uncertainty may be reduced as respondents would be able to access both possibilities—being a passenger and a pedestrian. For instance, being in the car and benefiting from the passenger-protective algorithm does not expose the respondents to the greater danger/risk of other cars employing the same algorithm when they are not in the car (e.g., as pedestrians). In addition to the partial information in the scenarios presented by Bonnefon et al. (2016), we further argue that the questions the authors claim to produce conflicting results in studies 3, 4, and 6 do not fully account for the willingness to buy an AV car. In the experiments participants were presented with a moral scenario where an AV can either be programmed to be utilitarian, passenger-protective, or select either option at random. This scenario was followed by questions, one of which required the participants to rate their relative willingness to buy an AV for themselves—“How would you rate your relative willingness of having an AV with each of these algorithms?.” The results revealed that participants preferred to purchase the passenger-protective AV, which once again conflicted with their general preference for utilitarian AV. Due to this conflict, our ongoing research aims to comprehensively understand utilitarian behavior by providing respondents with two moral questions regarding their willingness to purchase an AV, and their willingness for other people to purchase an AV: “Please rate how willing you would be to purchase an AV that is programmed with each of these algorithms” and “Please rate how willing you would be for other people to purchase an AV that is programmed with each of these algorithms.” It is plausible that full accessibility to moral tasks and questions reduces decision uncertainty and maximizes utility in moral decision-making with AVs. In our proposal, we argue that utility maximization can be increased by enabling participants to imagine themselves as not only as a passenger of an AV, but also as a pedestrian, and measure their judgments appropriately. RM drafted and revised the manuscript. PK (corresponding author) initiated and revised the general commentary. IK, AC, VB, and PV provided feedback and suggestions. All authors approved the final version of the manuscript for submission. The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Bentham, J. (1970). An Introduction to the Principles of Morals and Legislation. Darien, CT: Hafner (Original work published 1789). Google Scholar Bonnefon, J.-F., Shariff, A., and Rahwan, I. (2016). The social dilemma of autonomous vehicles. Science 352, 1573–1576. doi: 10.1126/science.aaf2654 PubMed Abstract | CrossRef Full Text | Google Scholar Dawes, R. M. (1980). Social dilemmas. Annu. Rev. Psychol. 31, 169–193. doi: 10.1146/annurev.ps.31.020180.001125 CrossRef Full Text | Google Scholar Greene, J. D., Sommerville, R. B., Nystrom, L. E., Darley, J. M., and Cohen, J. D. (2001). An fMRI investigation of emotional engagement in moral judgement. Science 293, 2105–2108. doi: 10.1126/science.1062872 CrossRef Full Text | Google Scholar Kollock, P. (1998). Social dilemmas: the anatomy of cooperation. Annu. Rev. Sociol. 24, 183–214. Google Scholar Kusev, P., van Schaik, P., Alzahrani, S., Lonigro, S., and Purser, H. (2016). Judging the morality of utilitarian actions: how poor utilitarian accessibility makes judges irrational. Psychon. Bull. Rev. 23, 1961–1967. doi: 10.3758/s13423-016-1029-2 PubMed Abstract | CrossRef Full Text | Google Scholar Thomson, J. J. (1985). The trolley problem. Yale Law J. 94, 1395–1415. doi: 10.2307/796133 CrossRef Full Text | Google Scholar Zhao, H., Dimovitz, K., Staveland, B., and Medsker, L. (2016). Responding to Challenges in the Design of Moral Autonomous Vehicles. The 2016 AAAI Fall Symposium Series: Cognitive Assistance in Government and Public Sector Applications, Technical Report FS-16-02, 169–173. Keywords: autonomous vehicles, utility theory, moral dilemmas, uncertainty, accessibility Citation: Martin R, Kusev I, Cooke AJ, Baranova V, Van Schaik P and Kusev P (2017) Commentary: The Social Dilemma of Autonomous Vehicles. Front. Psychol. 8:808. doi: 10.3389/fpsyg.2017.00808 Received: 28 January 2017; Accepted: 02 May 2017; Published: 24 May 2017. Edited by: Reviewed by: Copyright © 2017 Martin, Kusev, Cooke, Baranova, Van Schaik and Kusev. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. *Correspondence: Petko Kusev, [email protected]
Malte Schilling, , Katharina Muelling, Britta Wrede, Helge Ritter
Published: 14 May 2019
Frontiers in Neurorobotics, Volume 13; https://doi.org/10.3389/fnbot.2019.00016

Abstract:
Editorial on the Research TopicShared Autonomy—Learning of Joint Action and Human-Robot Collaboration Advancing the autonomy of artificial agents, such as robots, comes with an important challenge: how to shape the autonomy of an individual agent in such a manner that bringing several such agents together leads to a suitable fusion or sharing of (parts of) their individual autonomy spaces. This challenge has opened up important new perspectives on how to make robots more skillful, versatile, and easier to deploy for tasks and scenarios where robots can no longer act solitarily. The special issue aims at, first, providing a theoretical perspective to describe autonomy in collaborative settings, and, secondly, presenting novel results that highlight common traits and underlying dimensions and can guide further developments on adaptive interaction in teams of humans and robots. Accordingly, Shared Autonomy has to address the question on how to mediate autonomy between participating actors. This question can be approached through a distinction of different levels of autonomy and by defining characteristic dimensions describing collaborative scenarios (see Figure 1 and Schilling et al., 2016). In their review, Alonso and de la Puente identify transparency as one such characteristic which is understood as observability and predictability. They show that this translates well to interactive multi-agent scenarios and provides useful dimensions to account for interaction patterns. As predictability decreased in more complex tasks, systems were found to act more autonomously. They reasoned that this is due to limited observability which can be counteracted by subsuming low-level details in higher-level representations as are goals. Here, Gildert et al. continue with their review on joint action which highlights a shared context as a requirement for prediction. Furthermore, they point out the role of attention and how implicit and explicit communication can guide attention between agents in order to mediate interaction patterns. Both articles connect the issue of shared information in a further step toward the notion of trust in interactive settings. Figure 1. Distinguishing multiple levels of autonomy (shown on the left). Conceptually, these different levels characterize the freedom of decision making arising at different levels of an abstraction hierarchy. The notion of Shared Autonomy allows to analyze and design types of interaction patterns between human users and other agents (shown on the right). For details see Schilling et al. (2016). The special issue also provides examples that highlight current advances in human-robot collaboration and show that different levels of a representational hierarchy may call for different mechanisms and strategies: lower levels deal with the optimal selection of means to achieve immediate goals; autonomy on an intermediate level determines strategies that fulfill higher level purposes which are given on the highest level (as intentions). Importantly, studies on coordination with robots on a higher-level further stress that this can't be addressed solely on an abstract level, but always requires grounding in lower action levels. This further emphasizes the importance of a differentiated perspective on Shared Autonomy. For coordination on the lower level of immediate action, Trinh et al. provide a solution to navigation for multiple agents as a prototypical example for Shared Autonomy scenarios. Each agent preplans its own route to a target location which is represented as a static flow field. To avoid collisions, this representation is overlaid with a constantly updated dipole flow field of other agents' locations. The agents' behavior is mostly driven by the preplanned flow field, but the dipole flow field shapes the routes of the agents autonomously. Ewerton et al. directly address this change of control and autonomy. Their framework deals with visual and haptic feedback to users while learning motor skills. In their case users are intended to learn drawing Japanese characters. Probabilistic models of skills are learned following a reinforcement learning approach and the skill likelihood is used to regulate the amount of feedback. The system autonomously steps in when needed and starts to provide guidance through haptic feedback. The focus of Akkaladevi et al. is on how cooperation can be organized between multiple agents in a collaborative assembly process. On the lowest level, this entails learning basic actions. On a higher level, the system aims at putting these into a sequence. In addition, the system attempts to propose suitable actions that are chosen depending on background knowledge about users' capabilities, current context and current goals. Knowledge was learned using incremental reinforcement learning. The Interactive Shared Solution Shaping paradigm by Reardon et al. focuses further on an intermediate level of representation for negotiating a shared plan: while the autonomous agent has its own planning process, a human user provides feedback based on his expert domain knowledge. This was realized for planning the route of a surveillance robot. While the user doesn't need to know all the details about, e.g., collision avoidance, mediation of the planning process requires transparency between the robot system and the user. A form of interaction is needed that details current routes to the user and allows to influence the (re)planning process. Establishing such a form of fast and reliable communication is crucial for systems that should work autonomously, but at the same time contribute toward a user's intention and provide valuable information. In Shukla et al. such multimodal coordination is further analyzed and applied. They introduce the probabilistic Proactive Incremental Learning framework that learns to associate hand gestures with manipulation actions in a collaborative assembly task. There, communication is required to establish common ground and align mutual beliefs between robot and user which exploits information about gaze. The system anticipated users' behaviors and goals in order to proactively assist the user. In a study with non-roboticist users they found that their proactive system reduced the interaction effort and was mostly preferred, but further cautioned that this requires trust. In a further user study, Schulz et al. turn toward the question of how preferred interaction with robots is task dependent. They categorize forms of collaboration along two dimensions: first, distinguishing independent actions and joint interaction; secondly, distinguishing sequences where order is crucial versus not. This was assessed in a series of table-top scenarios in which a human collaborated with a robot to build a given design in a blocks world. They found that mostly autonomous actions of the robot are more efficient and preferred. The special issue brings together current work and trends in Shared Autonomy. Collaboration in teams of humans and robots shifts more and more toward higher level tasks that not only include low level coordination of immediate action. This requires coordination of complex plans between multiple agents that can be adapted at runtime. Therefore, each agent is tasked with autonomous control on different levels, but also has to respect the autonomy of other agents and therefore has to adjust its degree of autonomy. Shared Autonomy provides a useful perspective on the underlying interaction patterns, how these are adapted and what requirements this poses to systems, like establishing common ground, transparency and shared beliefs as well as goals, trust, and efficient forms of communication. MS prepared the figure and wrote the paper. HR, WB, KM, and BW discussed and wrote the paper. This work was supported by the Cluster of Excellence Cognitive Interaction Technology CITEC (EXC 277) at Bielefeld University, which is funded by the German Research Foundation (DFG). The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Schilling, M., Kopp, S., Wachsmuth, S., Wrede, B., Ritter, H., Brox, T., et al. (2016). “Towards a multidimensional perspective on shared autonomy,” in Proceedings of the AAAI Fall Symposium Series 2016 (Stanford, CA). Google Scholar Keywords: shared autonomy, human-robot interaction, learning, collaboration, interactive robots Citation: Schilling M, Burgard W, Muelling K, Wrede B and Ritter H (2019) Editorial: Shared Autonomy— Learning of Joint Action and Human-Robot Collaboration. Front. Neurorobot. 13:16. doi: 10.3389/fnbot.2019.00016 Received: 01 February 2019; Accepted: 09 April 2019; Published: 14 May 2019. Edited by: Reviewed by: Copyright © 2019 Schilling, Burgard, Muelling, Wrede and Ritter. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. *Correspondence: Malte Schilling, [email protected]
, Mohamed Ali Hadj Taieb, Mohamed Ben Aouicha, Grischa Fraumann, Christian Hauschke, Lambert Heller
Frontiers in Research Metrics and Analytics, Volume 6; https://doi.org/10.3389/frma.2021.694307

Abstract:
Fully structured semantic resources representing facts in the form of triples (i.e., knowledge graphs) have a major function in driving computer applications, particularly the ones related to biomedicine, to library and information science and to digital humanities (Haslhofer et al., 2018; Sargsyan et al., 2020). They can be easily processed using Application Programming Interfaces (APIs, like REST APIs) and query languages (mainly SPARQL) to assess the reference semantic information and to generate accurate and precise interpretations and predictions, particularly when the analyzed data is multifactorial and ever-changing such as the COVID-19 knowledge (Turki et al., 2021c), information about the laureates of Nobel Prize in Literature (Lebuda and Karwowski, 2016), and the findings of scholarly publications (Fathalla et al., 2017). In particular, the role of open knowledge graphs to facilitate scientific collaboration has been stressed against the backdrop of the COVID-19 pandemic (Anteghini et al., 2020; Colavizza et al., 2021; Turki et al., 2021a). Effectively, the information included in textual or semi-structured resources such as electronic health records, scholarly publications, encyclopedic entries, and citation indexes can be converted into fully structured Research Description Framework (RDF) triples and included in knowledge graphs and then processed in near real-time using computer methods to obtain evolving research outputs that are automatically updated as the knowledge graphs feeding them is regularly curated. These living research outputs include systematic reviews (Wang and Lo, 2021), clinical trials (Servant et al., 2014), scientometric studies (Nielsen et al., 2017), and epidemiological studies (Turki et al., 2021b). However, the construction of knowledge graphs is a complex effort including the recognition of scholarly publications related to the scope of the semantic resource (Turki, 2018), the retrieval of abbreviations and terms for every concept (Turki et al., 2021a), and the extraction and validation of semantic relations (Turki et al., 2018a). Many projects depend on advanced neural network-driven machine-learning techniques for applying these tasks as these methods contribute to higher quality (Asada et al., 2021; Fei et al., 2021). However, these techniques are considered as black boxes and cannot be debugged to identify the reasons behind returned false results and consequently to solve these limitations in a transparent way (Turki et al., 2021b). What is more, the quality of these techniques is considered imperfect in some cases, requiring more time to achieve the same results as specific well-defined algorithms (Turki et al., 2021b). Here, Bibliometric-Enhanced Information Retrieval (BIR) has evolved as a novel field that utilizes bibliographic metadata to efficiently drive the extraction and refinement of semantic data from scholarly publications (Cabanac et al., 2018). This field contributed to the development of many intuitive and explainable algorithms for knowledge engineering. On the one hand, this has been achieved through the restriction of the analysis of full texts to the publications including a particular value of a metadata to reveal the bibliographic settings where assessed algorithms perform well or bad (Safder and Hassan, 2019). On the other hand, this could be done thanks to the analysis of the bibliographic information using taxonomies like MeSH and Wikipedia Category Graph (Hadj Taieb et al., 2020) or using the probabilistic heuristics and constraints inferred from publications using statistical models including TF*IDF (Ramos, 2003) or extracted from knowledge graphs using inference engines, particularly HySpirit (Fuhr and Rölleke, 1998) and F-OWL (Zou et al., 2004). In this opinion article, we explain how each type of bibliographic metadata can provide useful insights to enhance the automatic enrichment and fact-checking of knowledge graphs from scholarly publications based on the outcomes of research efforts about BIR. In the following subsections, we provide an overview on bibliographic metadata that plays a pivotal role when employing BIR. We selected these types of metadata based on its frequent use in BIR and its wide availability. Title, abstract, controlled keywords, citation analysis, section title as well as further metadata of scholarly publications can be easily retrieved compared to full texts that are sometimes hidden behind paywalls. As such, these types provide the opportunity for better precision and recall for information retrieval. To start, title and abstract are two metadata elements that can enrich knowledge graphs. Even though titles and abstracts of scholarly publications are written in natural languages and are not semi-structured as the other types of bibliographic information, they give insights about the purpose and outcomes of research outputs in a concise way. That is why they can efficiently represent the topics and the format of research papers (Stotesbury, 2003; Letchford et al., 2015). Consequently, the time-consuming and demanding natural language processing of full texts is not required when a brief analysis of titles and abstracts can return required information for information retrieval purposes. As such, the availability of open abstracts as research data has been recently emphasized by the Initiative for Open Abstracts (I4OA) (Tay et al., 2020). The application of feature-based measures of sentences’ semantic similarity to compare the titles or abstracts of two scholarly publications can be efficient to identify whether the two papers describe similar topics or not (Hadj Taieb et al., 2019; Hadj Taieb et al., 2020) and this can serve to contextualize the co-citation and citation links between papers as well as to filter term co-occurrences for a more precise knowledge graph construction and refinement (Hadj Taieb et al., 2020). This is particularly true for the domain of STEM (Science, Technology, Engineering, and Mathematics) since it features, compared to arts and humanities as well as social sciences, an agreed-upon vocabulary which mostly addresses directly its subjects. Semantic similarity measures are driven by external knowledge resources like knowledge graphs and ontologies and can consequently compare brief texts with high accuracy and speed (Hadj Taieb et al., 2015) and full transparency (Turki et al., 2021b) by contrast to other advanced techniques applied to full texts, particularly deep learning, semantic embeddings (Sargsyan et al., 2020), TF-IDF1 (White, 2018), and Latent Dirichlet Allocation (Jeong et al., 2014). The consideration of the format of the titles and abstracts when applying information retrieval techniques can be an important factor for advancing the state of the art of the knowledge engineering field. The letter case of words in titles and abstracts can be useful for many information retrieval applications. For example, it can help identify scholarly abbreviations that are generally written in uppercase letters (Zhou et al., 2006) or to extract structured abstracts including uppercase section titles (Ripple et al., 2011). Such algorithms should be considered with care, particularly when the title or the abstract is fully written in uppercase letters such as in Telford et al. (1985). In this situation, case-sensitive algorithms should not be applied to uppercase titles and abstracts as this can alter the efficiency of the methods. Such systems cannot even be applied to the full text of a scholarly publication when both the title and the abstract are in uppercase letters. Consequently, such decisions can only be made by human reading of the title, abstract, and full text, if necessary. The restriction of several natural language processing algorithms to the titles and abstracts of scholarly publications can be associated with higher accuracy rates for methods. The usage of several patterns in titles and abstracts, particularly parentheses, can be less complicated in titles and abstracts than in full texts and this explains in part the higher accuracy of parenthetic abbreviation extraction from titles (Zhou et al., 2006). For instance, there are more situations where parentheses are used in full texts for explaining facts, mentioning in-text references and defining p-values for evaluating assumptions when parentheses are mainly used in titles for stating abbreviations and declaring chemical formulas (Zhou et al., 2006). This phenomenon should raise concerns about the application of information retrieval methods only tested on titles and abstracts to full texts on the one hand and detailed guidelines for deciding when the used methods should be restricted to titles and abstracts to obtain a better precision and recall for information retrieval. As a rule, richer metadata on publications are available than presented in the previous section. This includes contextual information such as content classifications, relationships to other documents as well as structure of content within an article. Regarding classification of the publication as a whole, controlled keywords are featured as terms from a reference terminology that are used to label scholarly publications in several bibliographic databases. Examples of these keywords are KeywordPlus attributed to Web of Science records (Zhang et al., 2016) and MeSH Keywords assigned to PubMed publications (Turki et al., 2018a). The advantage of using these keywords is that they allow the use of a unique term and not of synonyms to assign each concept to publications and allow to prevent the redundancy of many variants of the same term across the maintained citation index allowing a more precise data mining and knowledge engineering of the bibliographic database (Henry and McInnes 2017; Turki et al., 2018a). That is why the co-occurrence analysis of controlled keywords, particularly MeSH Keywords, is nowadays used in semantic relation extraction and validation and provides a high accuracy rate for such an action (Henry and McInnes, 2017). Concerning MeSH Keywords, the recognition of a semantic relation can be achieved through the identification of the compatibility of the qualifiers of two significantly co-occurring keywords (e.g., Sofosbuvir/therapeutic use and Hepatitis C/drug therapy), of the complementarity of the qualifier of a keyword with the class of the heading of another largely co-occurring keyword [e.g., Sofosbuvir/therapeutic use and Hepatitis C (disease)], or of the association of the classes of the headings of two mainly linked keywords [e.g., Sofosbuvir [drug] and Hepatitis C (disease)] (Turki et al., 2018a). Despite the fact that citations also have some shortcomings, they are currently recognized as major information in Scientometrics as they can provide important details about the impact of scholarly publications as well as the evolution of scientific outcomes (Zhai et al., 2018). That is why they can be useful for refining and enriching the outputs of information retrieval from research papers. As the majority of research publications is most likely to be cited by and co-cited with related papers dealing with the same topic, initially considered papers for constructing and validating a knowledge graph can be odd ones and should not be processed if they do not belong to the citation or co-citation network of the topic of the semantic resource (Turki, 2018). By contrast, the papers that have the best centrality in the citation or co-citation network of the field of the knowledge graph should be considered as reference resources that should drive the beginning and reasoning of the information retrieval algorithms as these publications are the main papers in the field upon which all other papers have been developed (Diallo et al., 2016). Controlled keywords and citations can be combined together to provide an added value to knowledge graph creation and validation from scholarly publications. The sentence including an inline citation to a work can be a key for enriching information about the citing paper as well as the cited one (Aljaber et al., 2011). The controlled keywords and the title analysis of the cited paper can be used to enrich the semantics of the citing sentences and recognize a hidden scientific relation or entity that has been discussed without being clearly stated (Aljaber et al., 2011; Hadj Taieb et al., 2020). The analysis of the inline citation using automatic named entity annotation and scientific relation embedding can reveal controlled vocabularies and relations that are not originally used to describe the cited paper in bibliographic databases (Aljaber et al., 2011). In another context, several types of semantic relations are available in particular sections of a scholarly publication (Turki et al., 2018a; Alexander and de Vries, 2021). For example, information about research funding for a given paper can only be found in specific parts (Alexander and de Vries, 2021). Alexander and de Vries (2021) note that the choice of an algorithm is important at the beginning of a research project. They use their algorithm to extract funding information from scholarly publications. The advantage of this text is that it is typically included under the section “Acknowledgments” or “Funding Information” in scholarly publications and adheres to certain writing standards, for instance, “This research is funded by’” followed by the name of the research funder, in some cases the funding program, and finally the grant number. Similarly, the section titles in narrative literature reviews provide an outlook about the information included in each part where a section entitled “Symptoms” in a review about Hepatitis C includes semantic relations about the symptoms of the described medical condition (Turki et al., 2018a). Subsequently, considering section titles during semantic relation extraction and validation can not only reduce the complexity of the recognition of the relation types but also minimize the time allocated for such a task by restricting the process to the sections that are expected to include the required relations. What other metadata elements of scholarly publications can be considered when it comes to building knowledge graphs? Scholarly publications have different levels of evidence according to their settings, the age, the type, the status, the research area, and the source title of a given output. All these factors may influence the significance of its findings to the research community (Burns et al., 2011). The age of a research paper can typically determine whether the information included in it is outdated or not as terms and abbreviations might change over years due to nomenclature updates (AlRyalat et al., 2018) and as several findings can be disproven after a time period thanks to advances in experimentation techniques and scientific reasoning (Arbesman, 2013). Although science is an ever-evolving enterprise, it is also based on certain classical literature, for example, the importance of the founders of academic disciplines, such as sociology. Consequently, scientists have to be mainly based on new publications to create better and updated knowledge graphs about their topics of interest and even to predict the evolution of the constructed knowledge graph in the next years (Choudhury et al., 2020). The type of a given publication can affect the amount and quality of information it includes. When letters only present a limited number of facts in a few pages (Turki et al., 2018b), reviews provide a detailed overview of the concepts and findings related to a given topic from the synthesis of many papers and are consequently more adequate as resources for scholarly information retrieval (Burns et al., 2011; Turki et al., 2018a). Acronyms used in scholarly literature can serve as a reliable goal of matching concepts in a research field. Many acronyms seem to be established terminology that is referred to frequently, in an unambiguous way. It has been shown that even for large corpuses of scientific papers from diverse fields automated disambiguation can be applied unsupervized and at scale (Charbonnier and Wartena 2018; Veyseh et al., 2021). The status of a research publication can be also important for ensuring the quality of the extraction of scientific knowledge. Although bibliographic databases like PubMed2 and features like Crossmark3 state whether a publication is a preprint or a partially or fully retracted paper, most of the projects for the creation and validation of knowledge graphs do not consider this factor when retrieving facts from research papers (Sargsyan et al., 2020). The matter with considering preprints in information retrieval is that these publications have not undergone peer review (Glasziou et al., 2020) and their outputs can be dynamically changed over months (Oikonomidi et al., 2020). That is why using them to generate structured information about a given topic can harm the quality of the created resource and make it less trustworthy. As for retracted publications, they are papers that have been proven to include false or fabricated claims, and were rejected by the scientific community and eliminated from their journal of publication for this reason. Although retractions are continuously cited for various reasons (e.g., lacking knowledge about the retractions), applying information retrieval techniques on them can let the users of the returned semantic data reuse scientifically doubtful findings and probably make wrong interpretations (Sotudeh et al., 2020; Soltani and Patini, 2020). The restriction of the set of considered publications according to their research areas as revealed in citation indexes or through the analysis of author keywords allows refining the creation and validation of knowledge graphs by eliminating outputs outside the scope of the developed resource (Salatino et al., 2020). This prevents the overlapping of concepts from different fields when they are represented by the same polysemous term and consequently eliminates noise from the generated database. To be precise, a polysemous term has various meanings. The consideration of the research venue of publications can be also efficient in this context. Further than the ability to analyze source titles using semantic similarity measures, among other techniques, to verify the topics of interest of journals and conferences (Hadj Taieb et al., 2015), metrics about the venues such as the journal impact factor and the number of citations can be used to filter the considered sources and only consider the most prestigious and reliable ones (Pal, 2021). While the journal impact factor has shortcomings, it can be a useful indicator in this context. In this opinion article, we showed the kinds of semantic information that can be revealed from each type of bibliographic metadata and that can be later used to strengthen information retrieval for knowledge graph construction and validation from scholarly publications. Given this, we invite the scientific community to collaborative projects considering bibliographic information when extracting domain information from scholarly publications for the creation and validation of trustworthy and precise semantic resources. As a future direction of this work, we suggest investigating how bibliographic metadata can enhance information retrieval algorithms using a series of experiments comparing the accuracy of methods processing full texts of scholarly publications with the one of bibliometric-enhanced information retrieval approaches. We consequently propose to study how bibliometric-enhanced information retrieval can enhance knowledge graph construction and validation as well as other interesting computational tasks such as predicting future scientific breakthroughs and major prize winners, natural language generation and translation of scholarly texts, and the automation of the creation and update of various kinds of research outputs. Researchers may also consider the adaptation of BIR algorithms to support the augmentation of university-level courses and evaluation quizzes with explanatory excerpts from scholarly outputs and to recommend scholarly publications to fight online misinformation. As well, we recommend building a framework for explainable artificial intelligence that returns explanations of the use of machine learning models for a given task based on what is currently available about the matter in research papers. As far as the availability of the data needed for BIR is concerned, it is to be hoped for the future that initiatives such as the I4OA mentioned above will gain momentum and that the applicability and re-usability of bibliographic metadata for BIR will become easier. All authors listed have made a substantial, direct, and intellectual contribution to the work and approved it for publication. The work of HT, MA, and MBA is supported by the Ministry of Higher Education and Scientific Research in Tunisia (MoHESR) in the framework of Federated Research Project PRFCOV19-D1-P1. The contribution of GF to this article is supported by the Train2 Wind ITN that has received funding from the European Union’s Horizon 2020 research and innovation programme under Marie Skłodowska-Curie grant agreement number 861291. The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Significant parts of this research output have been generated through discussion with the organizers and participants of the International Workshop on Bibliometric-enhanced Information Retrieval at the European Conference on Information Retrieval in 2020 and 2021 (Cabanac et al., 2020). 1TF-IDF: term frequency–inverse document frequency. 2Further details can be found in the PubMed User Guide at https://pubmed.ncbi.nlm.nih.gov/help/. For PubMed, information about the status of a research publication can be automatically retrieved using Biopython (Chapman and Chang, 2000). 3Crossmark is a crowdsourcing tool and database developed by CrossRef Organization that allows to annotate the types and statuses of scholarly publications (Meyer, 2011). Alexander, D., and de Vries, A. P. (2021). “This Research Is Funded by: Named Entity Recognition of Financial Information in Research Papers.,” in Proceedings of the 11th International Workshop on Bibliometric-enhanced Information Retrieval co-located with 43rd European Conference on Information Retrieval (ECIR 2021), Lucca, Italy, April 1, 2021 (Lucca, Italy: CEUR Workshop Proceedings), 102–110. Google Scholar Aljaber, B., Martinez, D., Stokes, N., and Bailey, J. (2011). Improving MeSH Classification of Biomedical Articles Using Citation Contexts. J. Biomed. Inform. 44 (5), 881–896. doi:10.1016/j.jbi.2011.05.007 PubMed Abstract | CrossRef Full Text | Google Scholar AlRyalat, S. A., Rawashdeh, K., El khatib, O., Yasin, A., Alqadi, F., Saleh, N., et al. (2018). The Change from an Eponym to a Representative Name: Wegener to Granulomatosis with Polyangiitis. Scientometrics 117 (3), 2077–2089. doi:10.1007/s11192-018-2951-z CrossRef Full Text | Google Scholar Anteghini, M., D’Souza, J., Martins. dos. Santos, V. A. P., and Auer, S. (2020). “Representing Semantified Biological Assays in the Open Research Knowledge Graph,” in International Conference on Asian Digital Libraries, Kyoto, Japan, November 30–December 1, 2020 (Cham: Springer), 89–98. doi:10.1007/978-3-030-64452-9_8 CrossRef Full Text | Google Scholar Arbesman, S. (2013). The Half-Life of Facts: Why Everything We Know Has an Expiration Date (New York: Penguin). ISBN:9781101595299. Asada, M., Miwa, M., and Sasaki, Y. (2021). Using Drug Descriptions and Molecular Structures for Drug-Drug Interaction Extraction from Literature. Bioinformatics. doi:10.1093/bioinformatics/btaa907 CrossRef Full Text | Google Scholar Burns, P. B., Rohrich, R. J., and Chung, K. C. (2011). The Levels of Evidence and Their Role in Evidence-Based Medicine. Plast. Reconstr. Surg. 128 (1), 305–310. doi:10.1097/PRS.0b013e318219c171 PubMed Abstract | CrossRef Full Text | Google Scholar Cabanac, G., Frommholz, I., and Mayr, P. (2020). “Bibliometric-Enhanced Information Retrieval 10th Anniversary Workshop Edition,” in European Conference on Information Retrieval, Lisbon, Portugal, April 14–17, 2020 (Cham: Springer), 641–647. doi:10.1007/978-3-030-45442-5_85 CrossRef Full Text | Google Scholar Cabanac, G., Frommholz, I., and Mayr, P. (2018). Bibliometric-enhanced Information Retrieval: Preface. Scientometrics 116 (2), 1225–1227. doi:10.1007/s11192-018-2861-0 CrossRef Full Text | Google Scholar Chapman, B., and Chang, J. (2000). Biopython. SIGBIO Newsl. 20 (2), 15–19. doi:10.1145/360262.360268 CrossRef Full Text | Google Scholar Charbonnier, J., and Wartena, C. (2018). “Using Word Embeddings for Unsupervised Acronym Disambiguation,” in Proceedings of the 27th International Conference on Computational Linguistics, Santa Fe, NM, August 20–26, 2018 (International Committee on Computational Linguistics), 2610–2619. doi:10.25968/opus-1265 CrossRef Full Text | Google Scholar Choudhury, N., Faisal, F., and Khushi, M. (2020). Mining Temporal Evolution of Knowledge Graphs and Genealogical Features for Literature-Based Discovery Prediction. J. Informetrics 14 (3), 101057. doi:10.1016/j.joi.2020.101057 CrossRef Full Text | Google Scholar Colavizza, G., Costas, R., Traag, V. A., van Eck, N. J., van Leeuwen, T., and Waltman, L. (2021). A Scientometric Overview of CORD-19. PLoS One 16 (1), e0244839. doi:10.1371/journal.pone.0244839 CrossRef Full Text | Google Scholar Diallo, S. Y., Lynch, C. J., Gore, R., and Padilla, J. J. (2016). Identifying Key Papers within a Journal via Network Centrality Measures. Scientometrics 107 (3), 1005–1020. doi:10.1007/s11192-016-1891-8 PubMed Abstract | CrossRef Full Text | Google Scholar Fathalla, S., Vahdati, S., Auer, S., and Lange, C. (2017). “Towards a Knowledge Graph Representing Research Findings by Semantifying Survey Articles,” in International Conference on Theory and Practice of Digital Libraries, Thessaloniki, Greece, September 18–21, 2017 (Cham: Springer), 315–327. doi:10.1007/978-3-319-67008-9_25 CrossRef Full Text | Google Scholar Fei, H., Zhang, Y., Ren, Y., and Ji, D. (2021). A Span-Graph Neural Model for Overlapping Entity Relation Extraction in Biomedical Texts. Bioinformatics. doi:10.1093/bioinformatics/btaa993 CrossRef Full Text | Google Scholar Fuhr, N., and Rölleke, T. (1998). “HySpirit - A Probabilistic Inference Engine for Hypermedia Retrieval in Large Databases,” in International Conference on Extending Database Technology, Valencia, Spain, March 23–27, 1998 (Berlin, Heidelberg: Springer), 24–38. doi:10.1007/BFb0100975 CrossRef Full Text | Google Scholar Glasziou, P. P., Sanders, S., and Hoffmann, T. (2020). Waste in Covid-19 Research. BMJ 369, m1847. doi:10.1136/bmj.m1847 PubMed Abstract | CrossRef Full Text | Google Scholar Hadj Taieb, M. A., Ben Aouicha, M., and Bourouis, Y. (2015). “Fm3s: Features-Based Measure of Sentences Semantic Similarity,” in International Conference on Hybrid Artificial Intelligence Systems, Bilbao, Spain, June 22–24, 2015 (Cham: Springer), 515–529. doi:10.1007/978-3-319-19644-2_43 CrossRef Full Text | Google Scholar Hadj Taieb, M. A., Ben Aouicha, M., and Turki, H. (2019). “Paper Co-citation Analysis Using Semantic Similarity Measures,” in International Conference on Intelligent Systems Design and Applications, December 3–5, 2019 (Cham: Springer), 264–277. doi:10.1007/978-3-030-49342-4_26 CrossRef Full Text | Google Scholar Hadj Taieb, M. A., Ben Aouicha, M., and Turki, H. (2020). Semantic-driven Bibliometric Techniques for Co-citation Analysis. His 16 (2), 111–125. doi:10.3233/HIS-200288 CrossRef Full Text | Google Scholar Haslhofer, B., Isaac, A., and Simon, R. (2018). Knowledge Graphs in the Libraries and Digital Humanities Domain Encyclopedia of Big Data Technologies. Cham: Springer, 1–8. doi:10.1007/978-3-319-63962-8_291-1 CrossRef Full Text Henry, S., and McInnes, B. T. (2017). Literature Based Discovery: Models, Methods, and Trends. J. Biomed. Inform. 74, 20–32. doi:10.1016/j.jbi.2017.08.011 PubMed Abstract | CrossRef Full Text | Google Scholar Jeong, Y. K., Song, M., and Ding, Y. (2014). Content-based Author Co-citation Analysis. J. Informetrics 8 (1), 197–211. doi:10.1016/j.joi.2013.12.001 CrossRef Full Text | Google Scholar Lebuda, I., and Karwowski, M. (2016). Written on the Writer's Face: Facial Width-To-Height Ratio Among Nominees and Laureates of the Nobel Prize in Literature. Creativity Res. J. 28 (2), 207–211. doi:10.1080/10400419.2016.1162572 CrossRef Full Text | Google Scholar Letchford, A., Moat, H. S., and Preis, T. (2015). The Advantage of Short Paper Titles. R. Soc. Open Sci. 2, 150266. doi:10.1098/rsos.150266 PubMed Abstract | CrossRef Full Text | Google Scholar Meyer, C. A. (2011). Distinguishing Published Scholarly Content with CrossMark. Learned Publishing 24 (2), 87–93. Distinguishing published scholarly content with CrossMark, Learned Publishing. doi:10.1087/20110202 CrossRef Full Text | Google Scholar Nielsen, F. Å., Mietchen, D., and Willighagen, E. (2017). “Scholia, Scientometrics and Wikidata,” in European Semantic Web Conference, Portorož, Slovenia, May 28–June 1, 2017 (Cham: Springer), 237–259. doi:10.1007/978-3-319-70407-4_36 CrossRef Full Text | Google Scholar Oikonomidi, T., Boutron, I., Boutron, I., Pierre, O., Cabanac, G., and Ravaud, P. (2020). Changes in Evidence for Studies Assessing Interventions for COVID-19 Reported in Preprints: Meta-Research Study. BMC Med. 18 (1), 402. doi:10.1186/s12916-020-01880-8 PubMed Abstract | CrossRef Full Text | Google Scholar Pal, J. K. (2021). Visualizing the Knowledge Outburst in Global Research on COVID-19. Scientometrics 126, 4173–4193. doi:10.1007/s11192-021-03912-3 CrossRef Full Text | Google Scholar Ramos, J. (2003). “Using Tf-Idf to Determine Word Relevance in Document Queries,” in Proceedings of the first instructional conference on machine learning, Washington, DC, August 21–24, 2003 (AAAI), 242, 29–48. Google Scholar Ripple, A. M., Mork, J. G., Knecht, L. S., and Humphreys, B. L. (2011). A Retrospective Cohort Study of Structured Abstracts in MEDLINE, 1992-2006. J. Med. Libr. Assoc. 99 (2), 160–163. doi:10.3163/1536-5050.99.2.009 PubMed Abstract | CrossRef Full Text | Google Scholar Safder, I., and Hassan, S.-U. (2019). Bibliometric-enhanced Information Retrieval: a Novel Deep Feature Engineering Approach for Algorithm Searching from Full-Text Publications. Scientometrics 119 (1), 257–277. doi:10.1007/s11192-019-03025-y CrossRef Full Text | Google Scholar Salatino, A. A., Thanapalasingam, T., Mannocci, A., Birukou, A., Osborne, F., and Motta, E. (2020). The Computer Science Ontology: A Comprehensive Automatically-Generated Taxonomy of Research Areas. Data Intelligence 2 (3), 379–416. doi:10.1162/dint_a_00055 CrossRef Full Text | Google Scholar Sargsyan, A., Kodamullil, A. T., Baksi, S., Darms, J., Madan, S., Gebel, S., et al. (2020). The COVID-19 Ontology. Bioinformatics 36 (24), 5703–5705. doi:10.1093/bioinformatics/btaa1057 CrossRef Full Text | Google Scholar Servant, N., Roméjon, J., Gestraud, P., La Rosa, P., Lucotte, G., Lair, S. v., et al. (2014). Bioinformatics for Precision Medicine in Oncology: Principles and Application to the SHIVA Clinical Trial. Front. Genet. 5, 152. doi:10.3389/fgene.2014.00152 PubMed Abstract | CrossRef Full Text | Google Scholar Soltani, P., and Patini, R. (2020). Retracted COVID-19 Articles: a Side-Effect of the Hot Race to Publication. Scientometrics 125 (1), 819–822. doi:10.1007/s11192-020-03661-9 PubMed Abstract | CrossRef Full Text | Google Scholar Sotudeh, H., Barahmand, N., Yousefi, Z., and Yaghtin, M. (2020). How Do Academia and Society React to Erroneous or Deceitful Claims? the Case of Retracted Articles' Recognition. J. Inf. Sci., 016555152094585. doi:10.1177/0165551520945853 CrossRef Full Text | Google Scholar Stotesbury, H. (2003). Evaluation in Research Article Abstracts in the Narrative and Hard Sciences. J. English Acad. Purposes 2 (4), 327–341. doi:10.1016/S1475-1585(03)00049-3 CrossRef Full Text | Google Scholar Tay, A., Kramer, B., and Waltman, L. (2020). Why Openly Available Abstracts Are Important – Overview of the Current State of Affairs. Available at: https://leidenmadtrics.nl/articles/why-openly-available-abstracts-are-important-overview-of-the-current-state-of-affairs (Accessed May 1, 2021). Leiden Madtrics. Google Scholar Telford, M., Mooi, R., and Ellers, O. (1985). A New Model of Podial deposit Feeding in the Sand Dollar, Mellita Quinqujesperforata (Leske): the Sieve Hypothesis Challenged. Biol. Bull. 169 (2), 431–448. doi:10.2307/1541493 PubMed Abstract | CrossRef Full Text | Google Scholar Turki, H. (2018). Citation Analysis Is Also Useful to Assess the Eligibility of Biomedical Research Works for Inclusion in Living Systematic Reviews. J. Clin. Epidemiol. 97, 124–125. doi:10.1016/j.jclinepi.2017.11.002 PubMed Abstract | CrossRef Full Text | Google Scholar Turki, H., Hadj Taieb, M. A., and Ben Aouicha, M. (2021a). Enhancing Filter-Based Parenthetic Abbreviation Extraction Methods. J. Am. Med. Inform. Assoc. 28 (3), 668–669. doi:10.1093/jamia/ocaa314 PubMed Abstract | CrossRef Full Text | Google Scholar Turki, H., Hadj Taieb, M. A., and Ben Aouicha, M. (2018a). MeSH Qualifiers, Publication Types and Relation Occurrence Frequency Are Also Useful for a Better Sentence-Level Extraction of Biomedical Relations. J. Biomed. Inform. 83, 217–218. doi:10.1016/j.jbi.2018.05.011 PubMed Abstract | CrossRef Full Text | Google Scholar Turki, H., Hadj Taieb, M. A., and Ben Aouicha, M. (2018b). The Value of Letters to the Editor. Scientometrics 117 (2), 1285–1287. doi:10.1007/s11192-018-2906-4 CrossRef Full Text | Google Scholar Turki, H., Hadj Taieb, M. A., Shafee, T., Lubiana, T., Jemielniak, D., Ben Aouicha, M., et al. (2021c). Representing COVID-19 Information in Collaborative Knowledge Graphs: the Case of Wikidata. Semantic Web J. Google Scholar Turki, H., Taieb, M. A. H., and Aouicha, M. B. (2021b). Developing Intuitive and Explainable Algorithms through Inspiration from Human Physiology and Computational Biology. Brief. Bioinform. doi:10.1093/bib/bbab081 CrossRef Full Text | Google Scholar Veyseh, A. P. B., Dernoncourt, F., Nguyen, T. H., Chang, W., and Celi, L. A. (2021). “Acronym Identification and Disambiguation Shared Tasks for Scientific Document Understanding,” in Proceedings of the 1st workshop on Scientific Document, February 8–9, 2021 (AAAI). Understanding at AAAI-21. Google Scholar Wang, L. L., and Lo, K. (2021). Text Mining Approaches for Dealing with the Rapidly Expanding Literature on COVID-19. Brief. Bioinform., 22 (2), 781–799. doi:10.1093/bib/bbaa296 PubMed Abstract | CrossRef Full Text | Google Scholar White, H. D. (2018). Bag of Works Retrieval: TF*IDF Weighting of Works Co-cited with a Seed. Int. J. Digit Libr. 19 (2), 139–149. doi:10.1007/s00799-017-0217-7 CrossRef Full Text | Google Scholar Zhai, Y., Ding, Y., and Wang, F. (2018). Measuring the Diffusion of an Innovation: a Citation Analysis. J. Assoc. Inf. Sci. Tech. 69 (3), 368–379. doi:10.1002/asi.23898 CrossRef Full Text | Google Scholar Zhang, J., Yu, Q., Zheng, F., Long, C., Lu, Z., and Duan, Z. (2016). Comparing Keywords Plus of WOS and Author Keywords: A Case Study of Patient Adherence Research. J. Assn Inf. Sci. Tec 67 (4), 967–972. doi:10.1002/asi.23437 CrossRef Full Text | Google Scholar Zhou, W., Torvik, V. I., and Smalheiser, N. R. (2006). ADAM: Another Database of Abbreviations in MEDLINE. Bioinformatics 22 (22), 2813–2818. doi:10.1093/bioinformatics/btl480 PubMed Abstract | CrossRef Full Text | Google Scholar Zou, Y., Finin, T., and Chen, H. (2004). “F-owl: An Inference Engine for Semantic Web,” in International Workshop on Formal Approaches to Agent-Based Systems, Greenbelt, MD, April 26–27, 2004 (Berlin, Heidelberg: Springer), 238–248. doi:10.1007/978-3-540-30960-4_16 CrossRef Full Text | Google Scholar Keywords: data mining and knowledge discovery, information retrieval and extraction, knowledge graph (ontologies), bibliometric-enhanced information retrieval, bibliographic metadata Citation: Turki H, Hadj Taieb MA, Ben Aouicha M, Fraumann G, Hauschke C and Heller L (2021) Enhancing Knowledge Graph Extraction and Validation From Scholarly Publications Using Bibliographic Metadata. Front. Res. Metr. Anal. 6:694307. doi: 10.3389/frma.2021.694307 Received: 12 April 2021; Accepted: 19 May 2021; Published: 28 May 2021. Edited by: Reviewed by: Copyright © 2021 Turki, Hadj Taieb, Ben Aouicha, Fraumann, Hauschke and Heller. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. *Correspondence: Houcemeddine Turki, [email protected]
, Phillip Hamrick
Published: 31 May 2021
Frontiers in Psychology, Volume 12; https://doi.org/10.3389/fpsyg.2021.655297

Abstract:
Emojis are a form of ideograms, consisting of icons intended to represent facial expressions, emotions, objects, or other symbols, most commonly used in technologies such as smartphones, tablets, and computers. Although the smiley face (a commonly used emoji) first appeared in the 1960s (Bai et al., 2019), emoji use has recently become ubiquitous, with more than six billion being sent daily (Evans, 2017). Despite being discounted by some as a degraded form of communication that is ruining language (e.g., Jones, 2015), there is substantial linguistic evidence that emoji are not a threat to natural language, but, rather, they are a useful augmentation of natural language that enhance its communicative capacities (Evans, 2017). This is particularly true because the types of written communication found in text messages, emails, Tweets, and other such technologies are often prone to misinterpretation without paralinguistic cues (e.g., gesture, intonation, facial expressions). Emojis can function as paralinguistic information (Tantawi and Rosson, 2019), and they can help disambiguate alternative interpretations, convey emotion, sarcasm, and other kinds of information normally only available in spoken face-to-face communication (Holtgraves and Robinson, 2020). Indeed, emoji are so powerful in their communicative conveyance that an entirely emoji-based message was responsible for a teenager's alleged terrorist threat in New York in 2015 (Evans, 2017). Thus, emojis are a popular and robust form of meaning-making. Emojis have also proven to be effective in studying cognitive phenomena without the influence of language. For example, Marengo et al. (2017) asked participants to respond to a brief Big Five Personality inventory and to 91 emojis drawn from the Apple Color Emoji fontset. Findings suggest that of the 91 emojis presented participant responses to 36 were correlated with their responses to three of the five personality traits. The responses to emojis were most related to emotions and affective processing. These researchers also empirically developed an emoji based instrument to assess depressive symptoms (Marengo et al., 2019). Marengo et al. (2019) conducted two studies to develop the emoji-based depression assessment. In the first study they asked young adults to indicate if each of the 36 emojis presented represented a way they felt during much of the past week. They also asked participants to complete a 10-item depression inventory. The association between the emojis and the depression measure items were calculated and the emoji with the 10 strongest associations with depression inventory items were tested for convergent validity and regression analyses allowed for accurate detection of depressive symptoms using the 10-item emoji scale. The results of the studies by Marengo and colleagues demonstrate the utility of emojis beyond language, with opportunities to create and study language-free measures of various cognitive phenomena, including personality, depressive symptoms, and perhaps other individual differences. One possible way to increase the validity and robustness of such findings would be to use established norms regarding the interpretation of the emojis, thus allowing for experimental hypothesis testing regarding the relationships between emoji interpretation and other psychological constructs. Despite their ubiquity and communicative power, the cognitive mechanisms supporting emoji comprehension and use remain elusive, although some research is starting to shed light on their role in semantic processing. For example, Weissman and Tanner (2018) examined whether emoji could induce language-like semantic processing. Event-related potentials were collected while participants read sentences like “The cake she made was terrible.” These sentences were followed by emoji that matched (), mismatched (), or that indicated irony or sarcasm (). Relative to those control trials, irony and sarcasm emojis elicited P600 and P200 event-related potentials, similar to the potentials elicited by purely verbal irony, suggesting that emoji and language might be processed similarly, at least in the case of irony and sarcasm. Emojis appear to also enhance regular language processing. Chatzichristos et al. (2020) found that emojis (compared to pseudoword controls) trigger more complex processing of the words they are paired with. Emojis also appear to facilitate meaning comprehension when the meaning of an utterance is indirect (Holtgraves and Robinson, 2020). Not only can emojis facilitate comprehension of language, but they also can aid in disambiguating other emojis. For example, adding a wink emoji to a message with food emojis that are not associated with sexual euphemisms can lead those same food emojis to be interpreted in a sexual way (Weissman, 2019). These findings make all the more sense in light of a recent study by Gantiva et al. (2020), who found that emoji faces elicited similar neural responses to human faces, suggesting again that emojis can provide paralinguistic information that is typically available in face-to-face, but not in text-based, communication. Emoji interpretation has also been studied without surrounding linguistic context. For example, Miller et al. (2017) examined definitions and sentiment ratings of ambiguous emojis, finding that emoji interpretation in context was not significantly less ambiguous than when they were interpreted standalone. There are at least two important caveats in much of the available research on how emojis are interpreted. First, the number of unique emojis used as stimuli has been restricted to a small pool of possible stimuli. Second, the “meanings” of the emojis used in these studies have often been assigned based on the researcher's intuition, rather than on the basis of data from norming studies. This may be particularly problematic in the case of emojis, which are very popular among the young adults who generally serve as participants in research. Given that research has shown differences between older and younger adults in emoji use (Hsiao and Hsieh, 2014) and interpretation (Gallud et al., 2018), it seems important for researchers to have a better sense of how young adults understand and interpret emojis before they begin designing their stimuli. Some researchers have attempted to gather such normative data. For example, Novak et al. (2015) used natural language processing tools to develop a publicly available sentiment association inventory for emojis based on the distributions of words that co-occurred with emojis in the social media platform Twitter. However, it is not clear how well such an inventory maps onto actual human ratings. Human ratings of emojis have been elicited in previous research (e.g., Miller et al., 2017), but these data have not been made publicly available to our knowledge. One set of normative data regarding emojis available to the public was produced by Rodrigues et al. (2018). Rodrigues et al. presented 258 stimuli (85 emoticons and 153 emojis) from the Lisbon Emoji and Emoticon Database to more than 500 Portuguese participants. Each participant rated 20 random stimuli on seven dimensions (e.g., valence, concreteness, meaningfulness). The results of the analysis indicated that emojis were found to have more aesthetic appeal and to be more meaningful than emoticons. Germaine to the current study, results of their meaning analysis suggest that intending meaning of the use of an emoji and the interpretation of the emoji are not always perfectly correlated. We believe that the Rodrigues et al. norms are useful for researchers as a set of norms however, we hope to provide a more comprehensive understanding of the way in which common emojis are interpreted. Our study also contributes to the literature as we conducted our study with English speaking participant using English responses, whereas the participants in the Rodrigues et al. (2018) were native Portuguese speakers. Given the growing interest in understanding the effects of emojis in semantic processing and given that there appears to be little or no publicly available data on how emojis are interpreted by the kinds of young adults that commonly participate in research, the aim of this study was to obtain data on how emojis are interpreted by young adults and to make those data and some code for how to parse them and use them for future research publicly available. These emoji interpretations could be, for example, (i) used in developing controlled experiments, (ii) compared with their meaning in context, (iii) can act as a baseline for computational accounts of emoji in meaning, and so on. To that end, this study provides a data set of 105 emoji from 129 participants. Sentiment (emotional valence) data and (multi)word associations for each emoji were elicited. Subject-level data on their use of emojis and text messaging behaviors was also elicited. The data are publicly available in their raw form and sample code for data manipulation, cleaning, and analyses are also available via the Open Science Framework at https://osf.io/za65c/. One hundred twenty-nine undergraduate psychology students at a large Midwestern University participated in the study for course credit. Eighty-eight participants identified as female and 41 as male. Participants were recruited from the university's Psychology Participant Pool via an online recruiting system. The age of participants ranged from 18 to 60 (M = 20, SD = 4.62). The university's Institutional Review Board approved the design of the study. Written consent was not collected as participants were ensured of the anonymity of the study and no identifying information was collected. 105 emojis were selected from the Apple® iPhone list of emojis as presented on the Emojipedia website (Emojis were selected based on their appearance in order in Apple's list of emojis (the emojis are viewable in the Open Science Framework https://osf.io/za65c/). The aim of the study was to present usable public data from a homogeneous pool of stimuli (e.g., controlled for size, coloration, etc.) from the most popular platform available. In this case, we chose emojis from Apple OS given available data on its popularity among college-aged people (e.g., among college-aged people in the USA, the Apple iPhone accounts from the majority of smartphone brand use, at ~40% of users according to Statista). That said, despite our goal of homogeneity of the pool of emojis, although we chose emojis from the Apple OS, the basic emoji symbols are virtually the same on iOS and Android as approved by the Unicode Consortium. The Apple and Google designers do create different looks for each icon and the names of the emojis are typically not representative of the emotions evoked, but are descriptive of the emoji itself (e.g., face with rolling eyes). In cases of emojis representing hand gestures, we randomly chose one example from the multiple skin tone representations. The study was programmed in E-Prime 3.0® and delivered to participants via the participant recruitment system as an E-Prime Go executable file. After scheduling participation in the study via the online participant recruitment system, participants downloaded the E-Prime Go file to their personal computer. Participants were informed that the study was designed to acquire individuals' interpretation of common emojis. The instructions informed participants that they would be shown emojis one at a time on their computer screen and asked to describe what the emoji was meant to represent and rate how strongly positive or negative the emotion related to the emoji occurred to them. Following the instructions, the participants were asked how often they send text messages. Participants replied using the following scale: 1 = 20 or more times per day. 2 = 10 or more times per day. 3 = 2 or more times per day. 4 = I rarely text. 5 = I never text. They were also asked to indicate how often they use emojis. They replied using the following scale: 1 = I include an emoji in almost every message I send. 2 = I use emojis in some messages. 3 = I rarely use emojis in my messages. 4 = I never use emojis in my messages. The 105 emojis appeared one at a time at the center of the computer screen. Above each emoji was the prompt “What word or phrase comes to mind with this emoji?” Participants typed in their response in a window appearing below the emoji and used the Enter key to complete their response. Following their response, the emoji remained in the center of the screen and the prompt “How strong is the emotion related to this emoji?” appeared above the emoji and a 5-point Likert-type scale ranging from 1-very negative to 5-very positive appeared below the emoji. Participants recorded their response using the numbers on their keyboard and the next emoji appeared. This procedure continued until participants responded to all 105 emojis. Participants' data was automatically saved to the desktop of their computer. Participants emailed the zip file containing the data to an email dedicated for data collection. In this section, we briefly describe some of the statistical characteristics of participants' responses, and we give some examples of the kinds of analyses that can be conducted with these data. Table 1 presents text use, emoji use, emotional valence of emoji, and number of words produced per emoji by self-reported sex. As expected, there was little variation among the participants in their text and emoji use. Participants generally reported very frequent use of text messaging (M = 1.44, SD = 0.77, SE = 0.07, range = 1–4). Similarly, participants reported frequently using emojis (M = 1.93, SD = 0.55, SE = 0.05, range = 1–3). No participants reported that they never text or never use emojis. The median emotional valence rating on a scale from 1 (very negative) to 5 (very positive) was computed for each emoji. The median of these values was 3, with some degree of variability across all emojis (SD = 0.95, SE = 0.09), and all possible valence values were used (range 1–5) (See Figure 1). Table 1. Text use, emoji use, emotional valence of emoji, and number of words produced per emoji by self-reported sex. Figure 1. Histograms of text use (A), emoji use (B), and emotional valence (C). Participants produced an average of 1.51 individual words (SD = 0.30, SE = 0.03, min = 1.22, max = 3.03) for each emoji, sometimes as strings of single words, sometimes expressing a more complex interpretation (e.g., “crying laughing,” “mind blown”). Figure 2 presents the 20 most produced words overall. Importantly, the raw data made available include participants' original responses, and researchers wishing to use these data can either work with the individual words, participants' multi-word responses, or both. Code for converting the data to a one-word-per-row are included in the R code also made available in this data report. Figure 2. The 20 most frequently produced individual words overall. The R code included with this data set also provides code that allows researchers to merge the emoji data with other data frames. Word-level data (e.g., word frequency, contextual diversity, orthographic neighborhood density, etc.) from existing databases (e.g., the English Lexicon Project, MRC Psycholinguistic Database, etc.) can be easily imported and entered into subsequent analyses. A potentially fruitful avenue for future research on how emojis are interpreted and how they influence meaning comprehension could be to study psycholinguistic properties of the words they have elicited in this dataset. Each emoji in our data set can produce a variety of word associations (e.g., see Figure 3), and those could be explored along several lines. For example, do certain emoji elicit more concrete or abstract word associations? It also may prove interesting to see the range of different emoji associated with a single word associate (Table 2). Figure 3. Word associations produced in response to (angry face). Note that the word responses here have not been filtered to remove stopwords (e.g., “to”) nor have they been modified (e.g., to correct spelling errors). Table 2. Emoji stimuli that elicited the word associate “happy” as a response at least 10 times across subjects (this is an arbitrary boundary used for simplicity of presentation). We also believe that another illuminating avenue for research will be to examine whether the semantic similarity of any two given emojis is comparable when using purely distributional information (e.g., Novak et al., 2015) compared with the emoji-(multi)word associations produced by our participants. In order to do so, one must convert the qualitative (multi)word responses into quantitative data. One way to do that is by using word vectors, which themselves are also learned distributionally. In the supplemental R code, we use pre-trained word vectors1 from a prominent distributional model of semantic memory, the Bound Encoding of the Aggregate Language Environment model (BEAGLE; Jones and Mewhort, 2007). The BEAGLE model produces word vectors that reflect both word context (e.g., semantic co-occurrence) and word order (e.g., syntax). These word vectors can then be compared (e.g., by computing their cosine) to determine their similarity, and by extension, the similarity of the words they represent. When a given emoji elicits one or more words, these vectors can be averaged to produce a composite (e.g., a “prototype” meaning for the emoji) or even compared against one another to determine how much internal consistency there is in the meaning of a given emoji. In the accompanying R code, we include the necessary code to import these vectors (vectors from other distributional models could also be used) along with code for computing cosine similarity between all average emoji vectors (we used a one-word-per-row format to do this, but it is possible to use participants' multi-word responses). These can be examined at a large scale, used for clustering emoji into groups, or for simply computing meaning similarity between individual pairs of emoji. For example, the cosine similarity between and , is very high (cos = 0.92), but is much less similar to (cos = 0.49). We acknowledge that there other publicly available data sets and analyses of emojis (e.g., Barbieri et al., 2016a,b), but our data set introduces publicly available human ratings that could act as human benchmarks, against which computational models such as can be tested. The aim of this study was to provide publicly available norms of interpretation for common emojis as well as some code to facilitate future research in rapidly growing areas of inquiry examining how emojis are interpreted in meaningful communication, how they interact with linguistic processing, and how computers might be able to process their meaning, just to name a few. The scale response for emotional valence makes these data easily comparable to other sentiment values for emoji that already exist (e.g., Novak et al., 2015). The open-ended word associations allow for individual differences in word choice and length of response. These (multi)word associations can be analyzed with respect to subject-level factors such as participant age and sex as well as text use and emoji use. They can also be analyzed at the item-level, examining variance in emotional valence and word association for each emoji. These data can also easily be merged with other data, be they statistical properties of the words (e.g., word frequency), semantic properties of the words (e.g., concreteness ratings, semantic vectors), or properties of the emojis themselves (e.g., emoji frequency). A follow-up step in this line of research could be to improve these data by including emoji from other platforms, from older participants, and from international populations, all of which would improve our understanding of the roles of these ubiquitous features of modern communication. The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/Supplementary Material. The studies involving human participants were reviewed and approved by Kent State University Institutional Review Board. Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements. CW designed and programmed the experiment, was responsible for the collection of data, and contributed to the manuscript writing and revisions. PH was the originator of the idea, organized and analyzed the data, and contributed to the writing and revisions of the manuscript. All authors contributed to the article and approved the submitted version. The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. We are grateful to Randall Jamieson for sharing the R code for making the pre-trained word vectors with us. The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpsyg.2021.655297/full#supplementary-material 1. ^We are grateful to Randall Jamieson for sharing the R code for making these vectors with us. Bai, Q., Dan, Q., Mu, Z., and Yang, M. (2019). A systematic review of emoji: current research and future perspectives. Front. Psychol. 10:2221. doi: 10.3389/fpsyg.2019.02221 PubMed Abstract | CrossRef Full Text | Google Scholar Barbieri, F., Kruszewski, G., Ronzano, F., and Saggion, H. (2016a). “How cosmopolitan are emojis? Exploring emojis usage and meaning over different languages with distributional semantics,” in Proceedings of the 24th ACM International Conference on Multimedia, 531–535. Google Scholar Barbieri, F., Ronzano, F., and Saggion, H. (2016b). “What does this emoji mean? a vector space skip-gram model for twitter emojis,” in Proceedings of the Tenth International Conference on Language Resources and Evaluation (LREC 2016), eds N. Calzolari, K. Choukri, T. Declerck T, et al., editors. Portorož; Paris: European Language Resources Association (ELRA), 3967–3972. Google Scholar Chatzichristos, C., Morante, M., Andreadis, N., Kofidis, E., Kopsinis, Y., and Theodoridis, S. (2020). Emojis influence autobiographical memory retrieval from reading words: an fMRI-based study. PLoS ONE 15:e0234104. doi: 10.1371/journal.pone.0234104 PubMed Abstract | CrossRef Full Text | Google Scholar Evans, V. (2017). The Emoji Code: The Linguistics Behind Smiley Faces and Scaredy Cats. New York, NY: Picador. Google Scholar Gallud, J., Fardoun, H., Andres, F., and Safa, N. (2018). “A study on how older people use emojis,” in Proceedings of the XIX Interactional Conference on Human Computer Interaction (New York, NY), 1–4. Google Scholar Gantiva, C., Sotaquirá, M., Araujo, A., and Cuervo, P. (2020). Cortical processing of human and emoji faces: an ERP analysis. Behav. Inform. Technol. 39, 935–943. doi: 10.1080/0144929X.2019.1632933 CrossRef Full Text | Google Scholar Holtgraves, T., and Robinson, C. (2020). Emoji can facilitate recognition of conveyed indirect meaning. PLoS ONE 15:e0232361. doi: 10.1371/journal.pone.0232361 PubMed Abstract | CrossRef Full Text | Google Scholar Hsiao, K., and Hsieh, P. (2014). “Age difference in recognition of emoticons,” in Proceedings of the International Conference on Human Interface and the Management of Information (Las Vegas, NV), 394–403. Google Scholar Jones, J. (2015, May 27). Emoji is dragging us back to the dark ages—and all we can do is smile. The Guardian. Available online at: https://www.theguardian.com/artanddesign/jonathanjonesblog/2015/may/27/emoji-language-dragging-us-back-to-the-dark-ages-yellow-smiley-face (accessed May 20, 2020). Google Scholar Jones, M. N., and Mewhort, D. (2007). Representing word meaning and order information in a composite holographic lexicon. Psychol. Rev. 114, 1–37. doi: 10.1037/0033-295X.114.1.1 PubMed Abstract | CrossRef Full Text | Google Scholar Marengo, D., Giannotta, F., and Settanni, M. (2017). Assessing personality using emoji: an exploratory study. Pers. Individ. Dif. 112, 74–78. doi: 10.1016/j.paid.2017.02.037 CrossRef Full Text | Google Scholar Marengo, D., Settanni, M., and Giannotta, F. (2019). Development and preliminary validation of an image-based instrument to assess depressive symptoms. Psychiatry Res. 279, 180–185. doi: 10.1016/j.psychres.2019.02.059 PubMed Abstract | CrossRef Full Text | Google Scholar Miller, H., Kluver, D., Thebault-Spieker, J., Terveen, L., and Hecht, B. (2017). “Understanding emoji ambiguity in context: the role of text in emoji-related miscommunication,” in Proceedings of the Eleventh International AAAI Conference on Web and Social Media (Montreal, QC), 152–161. Google Scholar Novak, P. K., Smailović, J., Sluban, B., and Mozetič, I. (2015). Sentiment of emojis. PLoS ONE 10:e0144296. doi: 10.1371/journal.pone.0144296 CrossRef Full Text | Google Scholar Rodrigues, D., Prada, M., Gaspar, R., Garrido, M. V., and Lopes, D. (2018). Lisbon Emoji and Emoticon Database (LEED): norms for emoji and emoticons in seven evaluative dimensions. Behav. Res. Methods 50, 392–405. doi: 10.3758/s13428-017-0878-6 PubMed Abstract | CrossRef Full Text | Google Scholar Tantawi, Y., and Rosson, M. B. (2019). “The paralinguistic function of emojis in Twitter communication.” in Proceedings of the 7th Conference on CMC and Social Media Corpora for the Humanities (Cergy), 68–72. Google Scholar Weissman, B. (2019). “Peaches and eggplants or…something else? The role of contex t in emoji interpretations,” in Proc. Linguist. Soc. Am. 4, 1–6. doi: 10.3765/plsa.v4i1.4533 CrossRef Full Text | Google Scholar Weissman, B., and Tanner, D. (2018). A strong wink between verbal and emoji-based irony: how the brain processes ironic emojis during language comprehension. PLoS ONE 13:e0201727. doi: 10.1371/journal.pone.0201727 PubMed Abstract | CrossRef Full Text | Google Scholar Keywords: emoji, emoji analysis, texting, emoji interpretation, emoji usage Citation: Was CA and Hamrick P (2021) What Did They Mean by That? Young Adults' Interpretations of 105 Common Emojis. Front. Psychol. 12:655297. doi: 10.3389/fpsyg.2021.655297 Received: 19 January 2021; Accepted: 03 May 2021; Published: 31 May 2021. Edited by: Reviewed by: Copyright © 2021 Was and Hamrick. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. *Correspondence: Christopher A. Was, [email protected]
, Alexia Zoumpoulaki, Parisa Eslambolchilar
Published: 17 May 2021
Frontiers in Virtual Reality, Volume 2; https://doi.org/10.3389/frvir.2021.673547

Abstract:
Cassandra Complex: from Greek mythology; someone whose valid warnings or concerns are disbelieved by others Recently, great steps have been taken in making virtual, augmented, and mixed reality (we refer to three realities as XR) technologies accessible to a broad and diverse end user audience. The sheer breadth of use cases for such technologies has grown, as it has been embedded into affordable, widely accessible, and on-the-go devices (e.g., iPhone) in combination with some popular intellectual property (e.g., Pokémon Go). However, with this increase has come recognition of several ethical issues attached to the widespread application of XR technologies in everyday lives. The XR domain raises similar concerns as the development and adoption of AI technologies, with the addition that it provides immersive experiences that blur the line of what is real and what is not, with consequences on human behavior and psychology (Javornik, 2016; Ramirez, 2019). It is easy to write off concerns with XR technology as unfounded or premature. However, the current state of the art in XR is capable of several use cases which we see as cause for concern: 1) XR can generate realistic holograms, thanks to advances in computer vision, of people. These hologram representations are lifelike and can be made to say or do things thanks to advances in deep fake technology where video footage of a person is generated in real time based on large data repositories of real captured footage (Westerlund, 2019). This can be used to promote disinformation. For example, a deepfake hologram portraying a movie celebrity sharing political propaganda which the celebrity themselves don’t endorse, targeting fans and spreading lies about the incumbent leader’s political opponents. The hologram could be made to harass or provoke viewers (Aliman and Kester, 2020), goading them into acting irrationally. This warrants ethical considerations when designing XR experiences for broadcasting and entertainment; 2) XR technology which can sense and interpret objects in the environment can be used to mask and/or delete recognized objects. This can be used to promote misleading and/or noncompetitive behavior in consumer goods marketing industries. For example, while a user is browsing an XR marketplace a soft drink manufacturer may identify a competitor’s can and make it look dented and/or undesirable, nudging the consumer to purchase their ‘superior’ looking product instead. In an XR environment, consumers have a more direct interaction with a product than in traditional broadcast based marketing, with XR providing powerful virtual affordances (Alcañiz et al., 2019) which can persuade consumers and their purchase intentions. When technology which can track our every move, and has knowledge of our preferences and desires, is given the power to make decisions on our behalf becomes widespread it may have unintended consequences (Neuhofer et al., 2020). Therefore, the use of XR in marketing should be subject to ethical considerations; 3) XR experiences may be so immersive that they distract from the user’s surroundings, opening them up to harm. For example, there have been several reports of Pokémon Go users being hit by passing vehicles as they play the game,1 completely immersed in the experience and unaware of what is happening around them in the real world. As these experiences revolve around storytelling, there are ethical responsibilities on the creators to ensure safe passage through the experience (Millard et al., 2019) for audiences and viewers, and as play takes place in real locations, one must consider the appropriateness of facilitating play in socio-historical or sacred locations (Carter and Egliston, 2020). Likewise, there are similar calls for standards in the design of experiences for educational purposes (Steele et al., 2020); and 4) XR technology can also be used to create realistic environments where though there may be no physical harm, certain experiences may expose participants to psychological trauma. Though there is no real threat in the environment, the participant perceives the virtual representation as such: it looks and 'feels' real. They may become overwhelmed with intense feelings of anxiety and fear as the graphical detail is staggering (Reichenberger et al., 2017; Lavoie et al., 2021; Slater et al., 2020). In this case, complex ethical situations arise when using XR technologies for therapy and research applications. These four use cases alone demonstrate the potential harm XR technologies may introduce for users, whether it be intentional or not, physical, or sociological. Social and political implications of emerging tech, for example social media sickening, are on the rise (Vaidhyanathan, 2018). The pace of emerging tools and technologies is so fast, as soon as we figure out what to do about one problem, a new one arises. Searching the Association for Computing Machinery (ACM) digital library reveals a growing trend in the area of XR and ethics that is nowhere close to slowing down (Figure 1). The point is: XR is following the same trend in publication outputs as AI. Given the bumps in the road that ethics and AI have observed in recent past, we note similar issues may begin to emerge soon in XR. Recent work has raised concerns over the practical utility of ethics documents written by governments, NGOs (Non-Governmental Organisations), and private sector agents. Schiff et al. cite several motivations extracted from a coding process over 80+ documents published between 2016 and 2020 regarding ethical approaches to AI (Schiff et al., 2020). They describe how motivations to publish documents can interact with one another; some agents may be motivated to act in a responsible manner, while others may be motivated to signal responsibility through publishing documents to increase their brand authority or take a leadership position for competitive advantage. XR is at risk of similar problems if all we do is publish policy documents with the aim to take positions on the global stage. FIGURE 1. Number of articles returned per year in the ACM Digital Library featuring the words mixed reality (ethics, mixed reality, virtual reality) vs. artificial intelligence (ethics, artificial intelligence, AI). As the application domain explodes, we must be vocal about the dangers to ethics and moral values in society. It may be necessary to impose a counsel for applied ethics upon developers and researchers: those who are creating and exploring XR technologies. As content creators shape applications and their use cases, it is naïve to pass the responsibility to policy makers: by this time, it is too late. We can no longer be reactive toward emerging problems in ethics of XR. We must strive for a proactive approach which goes beyond local policies and guidelines. If engineers and computer scientists are the ones to push the frontiers of XR technologies, we must accept our fallibility with grace and understand our own biases at play. We must work together with philosophers in ethics and governance to create a shared vision of what we want XR to be, together with industry, NGOs, and government to decide the best approach. This shared vision must consider not just the technical challenges to overcome in bringing immersive XR experiences to users, but how to do so in a way which is responsible and considers any and all hazardous consequences. In the current global political climate, where science and technology are sometimes viewed with hostility and mistrust2,3,4 (Caprettini and Voth, 2020), there may be a danger of modern Luddites arguing against AI, XR and similar technologies, afraid of adopting them even in cases where there may be obvious benefits, e.g., health, environment. We do not want to throw the baby out with the bathwater. Instead, we need to identify and address issues convincingly and thoroughly avoiding misleading the public. The literature has overwhelming suggestions on reforming these technologies and guidelines for design and development of algorithms, interfaces, and data collection. Although there are a few good practices adopted from industry such as establishing an ethics boards and accountability and privacy policies, these are self-regulated and self-governed. We lack a formula for how to approach the challenges we have identified as a community. We seem to act as policy reflexes: while there are many well-cited papers in the literature, proposals remain a series of ad-hoc insights or fragments of a larger whole. Having a formula and solid foundation will help to bring all these fragmented but invaluable efforts together. We argue for bringing advocates, policymakers, citizens, researchers, technologists, human-right activists, from different jurisdictions worldwide into the discussion to direct XR technologies for civil society. We need not look far for inspiration as promising steps toward reformation have been made in other domains. For example, there are invaluable lessons to be learned from responding to the climate crisis. For many decades, scientists rang the alarm bell about thinning ice sheets in the Arctic circles and increasing global temperatures. However, their international efforts were dismissed or ignored by the public and governments. As the overwhelming impact of global warming started affecting many populations around the world in the 1990s, the scientific community and activists were asked what they would propose instead. Thus the United Nations Framework Convention on Climate Change was defined, with most nations on earth agreeing to stabilize human induced greenhouse gas concentrations. In December 2015 official representatives from most other countries in the world gathered in Paris for the United Nations Climate Change Conference (or COP21), with legally binding consequences. In undertaking the Paris Agreement, governments agreed to work to limit temperature rise to well below 2 degrees Celsius. Agreements like this seek to form alliances across borders with like-minded nations in relation to climate change. A collective, united nations approach to ethical practice for mixed reality applications may prove beneficial and help to enact real impact and provide protective measures for participants, citizens, and consumers of XR research and products. We are currently seeing organized attempts to address the pressing ethical issues that have arisen from the wide adoption and development of AI solutions. In June 2019, the High-Level Expert Group on AI (AI HLEG) presented at the first European AI Assembly recommendations to guide trustworthy AI promoting “sustainability, growth and competitiveness, as well as inclusion–while empowering, benefiting and protecting human beings.”5 In August of the same year, the US National Institute of Standards and Technology released the Federal Engagement in Developing Technical Standards and Related Tools for AI.6 This document includes a set of actions the US federal government should take to protect public trust and confidence in AI as a priority, and emphasizes the importance of interdisciplinary research to increase understanding of issues around ethics and responsibility across society. Still, those actions lack the global approach that is needed in today’s interconnected world. It is encouraging to see early signs of global approach, for example the recent EU proposals7 containing regulations and guidelines for Excellence and Trust in AI. Agreeing on a framework for emerging technologies such as XR will not be as demanding as climate change on economic or political systems of nations who produce and export such technologies. However, to avoid the pitfalls associated with AI in the past, progress needs to happen on a global stage rather than localized approaches. The more frustrated academics become by not being heard the more tempted they may be to overstate, to provide even more information, and to use imperatives, all in all lowering communicative effectiveness. To avoid Cassandra’s fate, we need to appeal to people from all aspects of life via an international assembly. We need to shift discussions from ethics of XR to ethics for XR, and researchers across academia and industry must have effective communication with all stakeholders toward building a unified ethical framework for the development and deployment of XR technologies. DF, AZ, PE wrote the paper. DF, AZ, PE conducted ACM library search. AZ generated Figure 1. The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. 1https://time.com/4405221/pokemon-go-teen-hit-by-car/ 2https://www.independent.co.uk/life-style/gadgets-and-tech/news/google-artificial-intelligence-project-maven-ai-weaponise-military-protest-a8352751.html 3https://www.libertyhumanrights.org.uk/campaign/resist-facial-recognition/ 4https://futureoflife.org/ai-open-letter 5https://ec.europa.eu/digital-single-market/en/news/policy-and-investment-recommendations-trustworthy-artificial-intelligence 6https://www.nist.gov/system/files/documents/2019/08/10/ai_standards_fedengagement_plan_9aug2019.pdf 7https://digital-strategy.ec.europa.eu/en/library/proposal-regulation-laying-down-harmonised-rules-artificial-intelligence-artificial-intelligence Alcañiz, M., Bigné, E., and Guixeres, J. (2019). Virtual Reality in Marketing: A Framework, Review, and Research Agenda. Front. Psychol. 10, 1530. doi:10.3389/fpsyg.2019.01530 PubMed Abstract | CrossRef Full Text | Google Scholar Aliman, N.-M., and Kester, L. (2020). “Malicious Design in AIVR, Falsehood and Cybersecurity-oriented Immersive Defenses,” in 2020 IEEE International Conference on Artificial Intelligence and Virtual Reality (AIVR), 130–137. doi:10.1109/AIVR50618.2020.00031 CrossRef Full Text | Google Scholar Caprettini, B., and Voth, H.-J. (2020). Rage Against the Machines: Labor-Saving Technology and Unrest in Industrializing England. Ame. Econ. Rev. Insights 2, 305–320. doi:10.1257/aeri.20190385 CrossRef Full Text | Google Scholar Carter, M., and Egliston, B. (2020). Ethical Implications of Emerging Mixed Reality Technologies. Nanotechnology 14, 334–341. doi:10.25910/5ee2f9608ec4d CrossRef Full Text | Google Scholar Javornik, A. (2016). ‘It’s an illusion, but it looks real!’ Consumer Affective, Cognitive and Behavioural Responses to Augmented Reality Applications. J. Market. Manage. 32, 987–1011. doi:10.1080/0267257X.2016.1174726 CrossRef Full Text | Google Scholar Lavoie, R., Main, K., King, C., and King, D. (2021). Virtual Experience, Real Consequences: the Potential Negative Emotional Consequences of Virtual Reality Gameplay. Virtual Reality 25, 69–81. doi:10.1007/s10055-020-00440-y CrossRef Full Text | Google Scholar Millard, D. E., Hewitt, S., O'Hara, K., Packer, H., and Rogers, N. (2019). “The Unethical Future of Mixed Reality Storytelling,” in NHT ’19: Proceedings of the 8th International Workshop on Narrative and Hypertext (New York, NY, USA: Association for Computing Machinery), 5–8. doi:10.1145/3345511.3349283 CrossRef Full Text | Google Scholar Neuhofer, B., Magnus, B., and Celuch, K. (2020). The Impact of Artificial Intelligence on Event Experiences: a Scenario Technique Approach. Electron Markets 7, 133. doi:10.1007/s12525-020-00433-4 CrossRef Full Text | Google Scholar Ramirez, E. J. (2019). Ecological and Ethical Issues in Virtual Reality Research: A Call for Increased Scrutiny. Philos. Psychol. 32, 211–233. doi:10.1080/09515089.2018.1532073 CrossRef Full Text | Google Scholar Reichenberger, J., Porsch, S., Wittmann, J., Zimmermann, V., and Shiban, Y. (2017). Social Fear Conditioning Paradigm in Virtual Reality: Social vs. Electrical Aversive Conditioning. Front. Psychol. 8, 1979. doi:10.3389/fpsyg.2017.01979 PubMed Abstract | CrossRef Full Text | Google Scholar Schiff, D., Biddle, J., Borenstein, J., and Laas, K. (2020). “What’s Next for AI Ethics, Policy, and Governance? A Global Overview,” in AIES ’20: Proceedings of the AAAI/ACM Conference on AI, Ethics, and Society (New York, NY: Association for Computing Machinery), 153–158. doi:10.1145/3375627.3375804 CrossRef Full Text | Google Scholar Slater, M., Gonzalez-Liencres, C., Haggard, P., Vinkers, C., Gregory-Clarke, R., Jelley, S., et al. (2020). The Ethics of Realism in Virtual and Augmented Reality. Front. Virtual Real. 1, 113. doi:10.3389/frvir.2020.00001 CrossRef Full Text | Google Scholar Steele, P., Burleigh, C., Kroposki, M., Magabo, M., and Bailey, L. (2020). Ethical Considerations in Designing Virtual and Augmented Reality Products-Virtual and Augmented Reality Design with Students in Mind: Designers’ Perceptions. J. Educ. Technol. Syst. 49, 219–238. doi:10.1177/0047239520933858 CrossRef Full Text | Google Scholar Vaidhyanathan, S (2018). Antisocial Media: How Facebook Disconnects Us and Undermines Democracy. Oxford: Oxford University Press. Westerlund, M. (2019). The Emergence of Deepfake Technology: A Review. Technol. Innov. Manage. Rev. 9, 40–53. doi:10.22215/timreview/1282 CrossRef Full Text | Google Scholar Keywords: mixed reality, ethics, Cassandra Complex, virtual reality, artificial intelligence Citation: Finnegan DJ, Zoumpoulaki A and Eslambolchilar P (2021) Does Mixed Reality Have a Cassandra Complex?. Front. Virtual Real. 2:673547. doi: 10.3389/frvir.2021.673547 Received: 27 February 2021; Accepted: 30 April 2021; Published: 17 May 2021. Edited by: Reviewed by: Copyright © 2021 Finnegan, Zoumpoulaki and Eslambolchilar. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. *Correspondence: Daniel J. Finnegan, [email protected] These authors have contributed equally to this work and share first authorship
Published: 11 April 2014
Frontiers in Psychology, Volume 5; https://doi.org/10.3389/fpsyg.2014.00322

Abstract:
The research traditions of memory, reasoning, and categorization have largely developed separately. This is especially true for reasoning and categorization, where the former has focused on logic and probability rules and the latter on similarity processes. For example, classical rules of logic are often considered the basis for human reasoning (Evans et al., 1991) in tasks such as the Wason selection task (Wason, 1966), which requires participants to use deductive reasoning to solve a logic puzzle involving four cards. Reasoning models are typically developed in terms of hypotheses for how relevant rules should be combined and applied to reach conclusions from the relevant premises (Braine et al., 1995). By contrast, in categorization, the predominant theoretical traditions (i.e., prototype theory and exemplar theory) have involved a similarity process (Wills and Pothos, 2012, provide an overview). Such sharp distinctions between cognitive processes have started to break down. For example, Oaksford and Chater (1994) proposed a model of the Wason selection task based on information theory, rather than logical rules. Their model used information maximization in relation to uncertain hypotheses (cf. Anderson, 1991), which is an idea that could plausibly translate across other cognitive processes. Pothos (2010) showed how information maximization could apply to a learning task. Heit et al. (2012) argued that researchers should explore the connections between memory and reasoning. Tversky and Kahneman (1983), Shafir et al. (1990) proposed that when assessing the relative probability of statements about a hypothetical person, Linda, participants employ a process of similarity. Thus, the idea that similar or identical cognitive processes may underlie superficially disparate processes, like categorization and reasoning, is not new. What has been perhaps lacking is the development of specific models, which can be applied across different areas. Our purpose is to outline our ideas regarding such a model for probabilistic reasoning and similarity. We do so in the context of recent work with cognitive models based on quantum probability (QP) theory. Many people are familiar with quantum mechanics. What is perhaps less known is that the ingenious physicists who developed quantum mechanics also invented a new theory of probability, since classical probability (CP) theory was inconsistent with their bold new theory of the physical world. QP theory refers to the rules for assigning probabilities to events from quantum mechanics, without the physics. QP theory is potentially applicable to any area where there is a need to compute probabilities. The motivation to adopt QP theory is typically informed by whether the empirical situation of interest reflects some key properties of QP theory, such as incompatibility, interference, superposition, and entanglement. For example, when two possibilities are incompatible, this means that it is impossible to concurrently assign a truth-value to both. So, if we are certain about one possibility, then we are necessarily uncertain about the other. In CP theory it is always possible to create a complete joint probability distribution for all available alternatives. The quantum cognition research program is relatively new, but there have already been several notable applications, spanning conceptual combination (Aerts, 2009), perception (Atmanspacher et al., 2004), memory (Bruza et al., 2009), reasoning and decision-making (Busemeyer et al., 2011), and similarity (Pothos et al., 2013; for overviews see Busemeyer and Bruza, 2011 and Pothos and Busemeyer, 2013). We note that all these applications of QP theory in cognition have the form of standard cognitive models; they make claims regarding cognitive representations and processes, without any assumptions regarding the underlying neural substrate. In particular, quantum cognitive models do not require a quantum brain (the quantum brain hypothesis has been extremely controversial; Litt et al., 2006; Hammeroff, 2007). QP theory is a geometric approach to probability where different possibilities (or events or questions) are represented as subspaces, of varying dimensionality, in a multidimensional Hilbert space. Hilbert spaces are like vector spaces, but with some additional properties. The system of interest (e.g., the cognitive state of a participant in an experimental task) is a vector in the Hilbert space, called the state vector. Probabilities are determined by projecting the state vector onto different subspaces and computing the squared length of this projection. For example, consider Tversky and Kahneman's (1983) famous Linda problem, where participants read a story about a hypothetical person, Linda, and were asked to judge the likelihood of different features of Linda. We assume that the mental state vector corresponds to the representation of the information about Linda, after reading the story. As illustrated in Figure 1, the relevant Hilbert space includes subspaces for all relevant features of Linda, such as whether she is a feminist. Then, to extract the probability that a participant in the experiment will consider Linda to be a feminist, we project the state vector onto the feminist subspace, and compute the squared length of this projection. Generally, probability assignment in QP theory essentially involves a process of overlap between the cognitive state (modeled by the state vector) and different possibilities (modeled by subspaces). Thus, probability assignment in QP theory is a plausible candidate for a similarity process as well. Indeed, Sloman (1993) independently presented a model of induction, based on ideas closely resembling the formal properties of probabilistic computation in QP theory. Figure 1. The QP model for the conjunction fallacy. To calculate the probability that Linda is a bank teller, we project the state vector (shown in black) onto the bank teller subspace (shown in blue) and compute the squared length of this projection (shown in orange). To calculate the conjunction that Linda is both a feminist and a bank teller, we first project the state vector onto the feminist subspace (shown in red) and then project it onto the bank teller subspace and compute the squared length of this projection (shown in green). The process of projection in QP theory cognitive models is assumed to correspond to a process of thinking/evaluating the corresponding premise. However, for incompatible possibilities, it is impossible to identify a common projection to both relevant subspaces. This is equivalent to saying that it is impossible to concurrently assign a truth-value to incompatible observables. In order to compute a conjunction for incompatible observables (e.g., as needed for modeling the Linda experimental task), Busemeyer et al. (2011) postulated a process of sequential projection: the state vector is first projected onto the more likely predicate (feminism in the Linda example; cf. Gigerenzer and Goldstein, 1996) and then it is projected onto the less likely one (bank teller). The squared length of the resulting vector is the probability for the conjunction, according to QP theory. Note this number decomposes to the product of the probability of the marginal times the conditional, as one would expect from CP theory. Similar ideas of sequential projection have been used in other QP models to account for order effects in inference and causal reasoning (Trueblood and Busemeyer, 2011, 2012). The basic computation Pothos et al. (2013) proposed for the quantum similarity model is nearly identical. Suppose we are modeling the similarity of object A to object B. Similarity assessment is assumed to involve thinking about the first object and then the second, a process which can be modeled by projecting first onto the A subspace and then onto the B one, and computing the squared length of the resulting vector. Thus, the computation for assessing the similarity between A and B is exactly the same as for computing the conjunction between A and B, with two exceptions. The first relates to the order of projection. In decision-making tasks, we assume that the task does not typically impose any constraints on which premise is considered first. A corresponding choice has to be made and choosing the more likely premise first is a reasonable assumption. In similarity tasks, the similarity question is often phrased in a way that imposes a particular directionality (Tversky, 1977) and it is this directionality which constraints the order of projection. The second exception relates to how the initial mental state vector is determined. In decision-making applications, the state vector is plausibly determined by the information initially presented. In similarity, the state vector is set so that it does not bias the similarity judgment toward either of the compared objects. One may question these assumptions. However, QP theory is a mathematical theory of probability and we cannot expect psychological predictions to emerge without introducing some psychological assumptions too. The assumptions regarding projection order are exactly of this kind. It is worth noting that the quantum model for the conjunction fallacy predicts that if a person judges an unlikely event U before judging the conjunction of the unlikely event U and a likely event L, the effect is smaller than when the questions are in the opposite order, which has been confirmed experimentally by Stolarz-Fantino et al. (2003) and Gavanski and Roskos-Ewoldsen (1991). Overall, the decision-making model for the conjunction fallacy (Busemeyer et al., 2011) and the similarity model for Tversky's (1977) findings have similar forms. We think that this is a novel accomplishment, in that we provide a formal and testable expression of the idea that similarity and reasoning processes may involve the same or very similar cognitive mechanisms. Of course, the value of the modeling lies in the extent to which it can inform empirical prediction. Though speculative, we currently think there are a number of promising directions. For example, order effects in similarity judgments can arise because of differences in the degree of knowledge we have for one of the compared elements vs. the other (such differences can be related to the relative dimensionality of the corresponding subspaces). It is likewise possible that, in decision-making tasks, conjunctions in certain directions will be more natural than others. Tversky (1977) also reported a diagnosticity effect; where the similarity between two options resulted in a third, separate option being preferred in a matching task. Participants in this task were asked to decide which country was most similar to Austria. In one condition, the candidate choices were Sweden, Hungary, and Poland, and participants favored Sweden. However, when the candidate choices where changed to Sweden, Norway, and Hungary, participants preferred Hungary. Tversky explained the effect through a change in grouping–Eastern vs. Western Europe in the first condition and Nordic vs. non-Nordic in the second. This diagnosticity effect is analogous to a similarity effect in decision-making: when introducing an option similar to one of the existing options, this decreases the desirability of the similar options (cf. Trueblood et al., 2013). The quantum similarity model can capture the diagnosticity effect and the same mechanism has the potential to accommodate the corresponding result in decision-making. Likewise, the order of projections in the quantum similarity model allows for violations of the triangle inequality in similarity judgments and perhaps similar violations in decision-making can arise in the same way. The triangle inequality is a fundamental property of distance that must be obeyed by similarity measures based on simple functions of distance. In similarity judgments, the triangle inequality can be expressed as Similarity(A, B) > Similarity(A, C) + Similarity(C, B). Tversky (1977) reported a violation of the triangle inequality by having participants judge the similarity between three countries: Russia, Jamaica, and Cuba. He found that the similarity of Russia and Jamaica was smaller than the similarity of Russia and Cuba plus the similarity of Cuba and Jamaica. Is it meaningful to consider whether a general model of cognition could be based on QP principles? This is certainly an exciting proposition for researchers in the quantum cognition area. But a few qualifications need be noted. First, CP theory cognitive models can be and have been extremely successful too (e.g., Griffiths et al., 2010). A preference for the unique features of QP theory, against CP theory, can be motivated in situations where there is evidence of context effects, interference effects, and sequence effects. It is also possible that there will be empirical findings beyond both CP theory and QP theory. In fact, QP theory is a highly constrained framework and we already have simple, non-parametric tests, which can test for consistency with QP principles (Wang and Busemeyer, 2013). All of these possibilities provide exciting directions for future research. Although we did not discuss memory processes explicitly, there has been initial work in developing a QP model of memory recognition (Busemeyer and Trueblood, 2010). In sum, we think that the current quantum cognitive models provide an encouraging starting point in exploring the commonalities between memory, reasoning, and categorization, because of the simplicity of the basic ideas and their applicability across areas. The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Jennifer S. Trueblood was supported by NSF grant SES-1326275, Emmanuel M. Pothos by Leverhulme Trust grant RPG-2013-004, and Jerome R. Busemeyer by NSF grant ECCS – 1002188. Emmanuel M. Pothos and Jerome R. Busemeyer were supported by Air Force Office of Scientific Research (AFOSR), Air Force Material Command, USAF, grants FA 8655-13-1-3044 and FA 9550-12-1-0397 respectively. The U.S Government is authorized to reproduce and distribute reprints for Governmental purpose notwithstanding any copyright notation thereon. Aerts, D. (2009). Quantum structure in cognition. J. Math. Psychol. 53, 314–348. doi: 10.1016/j.jmp.2009.04.005 CrossRef Full Text Anderson, J. R. (1991). The adaptive nature of human categorization. Psychol. Rev. 98, 409–429. doi: 10.1037/0033-295X.98.3.409 CrossRef Full Text Atmanspacher, H., Filk, T., and Romer, H. (2004). Quantum zero features of bistable perception. Biol. Cybern. 90, 33–40. doi: 10.1007/s00422-003-0436-4 Pubmed Abstract | Pubmed Full Text | CrossRef Full Text Braine, M. D. S., O'Brien, D. P., Noveck, I. A., Samuels, M. C., Lea, B. L., Fisch, S. M., et al. (1995). Predicting intermediate and multiple conclusions in propositional logic inference problems: further evidence for a mental logic. J. Exp. Psychol. 124, 263–292. doi: 10.1037/0096-3445.124.3.263 CrossRef Full Text Bruza, P. D., Kitto, K., Nelson, D., and McEvoy, C. L. (2009). Is there something quantum-like about the human mental lexicon? J. Math. Psychol. 53, 362–377. doi: 10.1016/j.jmp.2009.04.004 Pubmed Abstract | Pubmed Full Text | CrossRef Full Text Busemeyer, J. R., and Bruza, P. (2011). Quantum Models of Cognition and Decision Making. Cambridge, UK: Cambridge University Press. Busemeyer, J. R., Pothos, E., Franco, R., and Trueblood, J. S. (2011) A quantum theoretical explanation for probability judgment “errors”. Psychol. Rev. 118, 193–218. doi: 10.1037/a0022542 Pubmed Abstract | Pubmed Full Text | CrossRef Full Text Busemeyer, J. R., and Trueblood, J. S. (2010). “Quantum model for conjoint recognition,” in AAAI Fall Symposium Series (Arlington, VA: Association for the Advancement of Artificial Intelligence). Evans, St B. T. J., Newstead, S. E., and Byrne, R. J. M. (1991). Human Reasoning: The Psychology of Deduction. Hove: Lawrence Erlbaum Associates. Gavanski, I., and Roskos-Ewoldsen, D. R. (1991). Representativeness and conjoint probability. J. Pers. Soc. Psychol. 61, 181–194. doi: 10.1037/0022-3514.61.2.181 Pubmed Abstract | Pubmed Full Text | CrossRef Full Text Gigerenzer, G., and Goldstein, D. (1996). Reasoning the fast and frugal way: models of bounded rationality. Psychol. Rev. 103, 650–669. doi: 10.1037/0033-295X.103.4.650 Pubmed Abstract | Pubmed Full Text | CrossRef Full Text Griffiths, T. L., Chater, N., Kemp, C., Perfors, A., and Tenenbaum, J. B. (2010). Probabilistic models of cognition: exploring representations and inductive biases. Trends Cogn. Sci. (Regul. Ed.) 14, 357–364. doi: 10.1016/j.tics.2010.05.004 Pubmed Abstract | Pubmed Full Text | CrossRef Full Text Hammeroff, S. R. (2007). The brain is both a neurocomputer and quantum computer. Cogn. Sci. 31, 1035–1045. doi: 10.1080/03640210701704004 Pubmed Abstract | Pubmed Full Text | CrossRef Full Text Heit, E., Rotello, C., and Hayes, B. (2012). “Relations between memory and reasoning,” in Psychology of Learning and Motivation, Vol. 57, ed B. H. Ross (San Diego, CA: Academic Press), 57–101. doi: 10.1016/B978-0-12-394293-7.00002-9 CrossRef Full Text Litt, A., Eliasmith, C., Kroon, F. W., Weinstein, S., and Thagard, P. (2006). Is the brain a quantum computer? Cogn. Sci. 30, 593–603. doi: 10.1207/s15516709cog0000_59 Pubmed Abstract | Pubmed Full Text | CrossRef Full Text Oaksford, M., and Chater, N. (1994). A rational analysis of the selection task as optimal data selection. Psychol. Rev. 101, 608–631. doi: 10.1037/0033-295X.101.4.608 CrossRef Full Text Pothos, E. M. (2010). An entropy model for artificial grammar learning. Front. Cogn. Sci. 1:16. doi: 10.3389/fpsyg.2010.00016 Pubmed Abstract | Pubmed Full Text | CrossRef Full Text Pothos, E. M., and Busemeyer, J. R. (2013). Can quantum probability provide a new direction for cognitive modeling? Behav. Brain Sci. 36, 255–327. doi: 10.1017/S0140525X12001525 Pubmed Abstract | Pubmed Full Text | CrossRef Full Text Pothos, E. M., Busemeyer, J. R., and Trueblood, J. S. (2013). A quantum geometric model of similarity. Psychol. Rev. 120, 679–696. doi: 10.1037/a0033142 Pubmed Abstract | Pubmed Full Text | CrossRef Full Text Shafir, E. B., Smith, E. E., and Osherson, D. N. (1990). Typicality and reasoning fallacies. Mem. Cogn. 18, 229–239. doi: 10.3758/BF03213877 Pubmed Abstract | Pubmed Full Text | CrossRef Full Text Sloman, S. A. (1993). Feature-based induction. Cogn. Psychol. 25, 231–280. doi: 10.1006/cogp.1993.1006 CrossRef Full Text Stolarz-Fantino, S., Fantino, E., Zizzo, D. J., and Wen, J. (2003). The conjunction effect: new evidence for robustness. Am. J. Psychol. 116, 15–34. doi: 10.2307/1423333 Pubmed Abstract | Pubmed Full Text | CrossRef Full Text Trueblood, J. S., Brown, S. D., Heathcote, A., and Busemeyer, J. R. (2013). Not just for consumers: context effects are fundamental to decision-making. Psychol. Sci. 24, 901–908. doi: 10.1177/0956797612464241 Pubmed Abstract | Pubmed Full Text | CrossRef Full Text Trueblood, J. S., and Busemeyer, J. R. (2011). A quantum probability account of order effects in inference. Cogn. Sci. 35, 1518–1552. doi: 10.1111/j.1551-6709.2011.01197.x Pubmed Abstract | Pubmed Full Text | CrossRef Full Text Trueblood, J. S., and Busemeyer, J. R. (2012). A quantum probability model of causal reasoning. Front. Cogn. Sci. 3:138. doi: 10.3389/fpsyg.2012.00138 Pubmed Abstract | Pubmed Full Text | CrossRef Full Text Tversky, A. (1977). Features of similarity. Psychol. Rev. 84, 327–352. doi: 10.1037/0033-295X.84.4.327 CrossRef Full Text Tversky, A., and Kahneman, D. (1983). Extensional versus intuitive reasoning: the conjunction fallacy in probability judgment. Psychol. Rev. 90, 293–315. doi: 10.1037/0033-295X.90.4.293 CrossRef Full Text Wang, Z., and Busemeyer, J. R. (2013). A quantum question order model supported by empirical tests of an a priori and precise prediction. Top. Cogn. Sci. 5, 689–710. doi: 10.1111/tops.12040 Pubmed Abstract | Pubmed Full Text | CrossRef Full Text Wason, P. C. (1966). “Reasoning,” in New Horizons in Psychology, ed B. M. Foss (Harmondsworth: Penguin), 135–151. Wills, A. J., and Pothos, E. M. (2012). On the adequacy of current empirical evaluations of formal models of categorization. Psychol. Bull. 138, 102–125. doi: 10.1037/a0025715 Pubmed Abstract | Pubmed Full Text | CrossRef Full Text Keywords: quantum probability theory, conjunction fallacy, similarity judgment, decision-making, classical probability theory Citation: Trueblood JS, Pothos EM and Busemeyer JR (2014) Quantum probability theory as a common framework for reasoning and similarity. Front. Psychol. 5:322. doi: 10.3389/fpsyg.2014.00322 Received: 03 March 2014; Accepted: 27 March 2014; Published online: 11 April 2014. Edited by: Reviewed by: Copyright © 2014 Trueblood, Pothos and Busemeyer. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. *Correspondence: [email protected]
Published: 8 July 2015
Frontiers in Psychology, Volume 6; https://doi.org/10.3389/fpsyg.2015.00929

Abstract:
The use of domestic service robots is becoming widespread. While in industrial settings robots are often used for specified tasks, the challenge in the case of robots put to domestic use is to afford human-robot collaboration in a variety of non-predefined and different daily tasks. Herein, we aim at identifying and understanding the conditions that will facilitate flexible collaboration between humans and robots. Past research of social and personality psychology was mainly focused on individual's self-regulation, defined as the ability to govern, or direct attention, resources, or action toward the realization of a particular goal (Higgins, 1989; Kruglanski et al., 2002). There is evidence that pursuing goals with the presence of others influences self-control (Fishbach and Trope, 2005), however only little is known on dyadic processes of self-regulation. Additionally, whereas research of goal pursuit in social psychology has mainly been associated with general processes of the structure and function of goals (Gollwitzer and Bargh, 1996; Carver and Scheier, 1998; Kruglanski et al., 2002; Fishbach and Ferguson, 2007; Custers and Aarts, 2010), human-robot interaction involves pragmatic interpersonal dilemmas such as how to coordinate human-robot activity and what knowledge should be shared between humans and robots over the course of action. To fill this gap, in what follows, we will define the unique characteristics of what we term as human-robot coupled self-regulation, which has the unique features of a dyadic asymmetric team aimed to increase the affordances of an individual in different activities. We will describe the unique characteristics of human-robot interaction and its special challenges toward goal pursuit. Our first assumption is that self-regulation of a human-robot couple could be conceptualized as a unique team configuration. A team is “a distinguishable set of two or more people who interact, dynamically, interdependently, and adaptively toward a common and valued goal/objective/mission, who have each been assigned specific roles or functions to perform, and who have a limited life-span of membership” (Salas et al., 1992, p. 4; Salas et al., 2010). Team members have differentiated responsibilities and roles (Cannon-Bowers et al., 1993). Therefore, essential for a team's successful performance is the understanding of the abilities and behaviors of its members that fit their experience and unique expertise for the task at hand. Because humans and robots differ in their level of agency (the capacity to act and do) and their level of experience (the capacity to feel and sense), (Gray and Wegner, 2012), we argue that their contribution to the team is not symmetric. Based on the reasoning that genuine authorship of an action or situation may not always be clear (Dijksterhuis et al., 2008), we suggest that defined requirements of person, robot, and situation are essential to reduce the expectation gap. Our perspective is that human-robot collaboration should be viewed in terms of functionality, to extend possibilities for the kinds of goals that humans want to pursue. These instrumental relations between a person and her tool, used to increase the fit between person and environment, are termed affordances (Gibson, 1979). Following this view, we argue that robots can be perceived as self-regulatory tools to increase affordances across different situations (Koole and Veenstra, 2015). Our instrumental relational approach enables flexibility in tuning the robot's level of responsiveness and dominance in human-robot social contexts. For example, whereas the human member of the team holds a fixed ownership position, the robot's level of dominance could vary by user demands, or depending on the situation. To understand the usefulness of this principle, let us take for example 80 year old Mrs. Brown. She is physically fragile, but it is important for her to maintain an independent life style. This is why she has “Rupert,” a multi-functional platform robot that serves as her aid. When she leaves the house she may want “Rupert” to lead and find the safest walking path to the store, thus she may set it to high dominance and responsiveness, in case she startles. At home, she may not desire high level of proactive care-taking and leave “Rupert” to be on call. Our second assumption is that human-robot coupled self-regulation is based on concrete rather than abstract level of agreement. Carrying out human-robot joint actions demands continuous coordination on at least five elements: (1) who takes part; (2) what is the role of each member; (3) what is the joint goal; (4) how does each team member contribute to the timing and synchronization; and (5) where the actions take place (Clark, 2005). To address this, the robot should identify where the focus of attention of the human is, to what degree the attention of the human is focused on team actions, and how to convey feedback. Similarly, the human needs to calibrate expectations from the robot, i.e., be invested in the robot's immediate action or approval of action, and how to respond to the robot's requests (Alami et al., 2005). Coupled self-regulation of goals requires agreement on goal setting and goal striving as two basic phases in goal pursuit (Gollwitzer and Oettingen, 2011). Whereas, robots may act automatically from initiation to completion of the task, humans' possible reflection on their performance may involve conscious awareness and create new representations of behavior, thus leading to communication gaps (Baumeister and Bargh, 2014). According to the action identification theory, a specific action can be verbally identified and interpreted from different levels of abstraction, ranging from low-level identities that specify how the action is performed, to high-level identities that signify why the action is performed. For instance, a person who “drinks water” can identify it as “holding a glass” (low level), or as “relieving thirst” (high level) (Vallacher and Wegner, 1987, 1989). This helps explain why different action identifications by human and robot may lead to dissimilar systems of goals and means of attainment (Kruglanski et al., 2002; Shah et al., 2002). To address these challenges, we suggest the use of multiple human-robot forms of communication to pursue the joint goal. Lohan et al. (2014) proposed a distinction between two kinds of actions: path-oriented and manner-oriented, that can be communicated via two different linguistic utterance styles. Whereas, in path-oriented utterances the goal is stressed, in manner-oriented utterances, the means of motion are emphasized (e.g., Talmy, 1991). In our example, Mrs. Brown and “Rupert” carry a recliner to the porch (Path-“let's move the chair to the porch” or Manner-“I want to read my book on the porch”). Suddenly the phone rings and Mrs. Brown wants to go and answer ((Path-“let me go get the phone” or Manner-“I need to answer this call”). “Rupert” must understand that the goal has changed and pause. Research indicates that professional and social interactions between team members can develop the team's social cognition (Klimoski and Mohammed, 1994). There is evidence that a team's fluent on-going communication regarding goal pursuit reduces the need for preexisting knowledge (Kozlowski and Bell, 2003). In social HRI, it is critical to generate many levels of interaction with the automation. Hence, the robot should always be present and aim to facilitate the goal, even if only to provide recommendations. In civil aviation, for example, communication is key especially if things turn out unexpectedly. In the Northwest 2009 incident in Minneapolis the automation had the capability, but was not designed to point out that the task was not performed as planned and that the pilots missed their destination. To borrow from our previous example, let us suppose Mrs. Brown wants to grab a pillow from the upper cabinet. The robot may not be able to reach so high, but it should continue to collaborate by providing feedback and advice; I cannot reach the uppermost cupboard (failure to complete task) but it is too dangerous for you to try to reach it on your own, if not urgent, perhaps we should call your son, or is there another pillow on a lower shelf? Much of human communication over goal pursuit is based on social cues (e.g., gestures, and mimicry) that automatically generate social judgment and behavior (Chartrand and Bargh, 1999; van Baaren et al., 2003; Leander et al., 2010). Similarly, translation of social cues to social signals leads to inference of human intentions by robotic agents (Fiore et al., 2013). The relevance of automatic embodied cues for joint goal pursuit was demonstrated in human-human and human-robot synchronicity, suggesting that physical synchronicity is associated with experience of responsiveness and empathy (Sebanz and Knoblich, 2009; Cohen et al., 2010; Paladino et al., 2010; Boucher et al., 2012; Hoffman et al., 2014). Embodied communication is not only “used” by robots, but integrated in them to support both the recognition of the human's behavior and the generation of their behavior. Research of social signal processing and modeling multimodal communication, suggests that social and behavioral cues may be detectable from a machine, hence perceivable. Likewise, models of behavior are integrated in a way that a robot exhibits a more natural behavior, aiming at a more successful interaction with the human (Pentland, 2007; Vinciarelli et al., 2012). However, despite emerging findings from the field of embodied cognition on the potential of physical and social cues as an alternative route for communication, it was also claimed that embodied cognition cues can lead to different patterns of activation across different contexts (Loersch and Payne, 2011), thus prediction of behavior may be difficult (Shalev, 2015). A possible way to address this limitation is to use robots in fixed context, where interpretation to human's embodied signals is less ambiguous. For example Loth et al. (2013), have demonstrated that bar staff responded to a set of two non-verbal signals. Foster (2014), indicated that robotic sensors can similarly detect and respond to these signals. Individuals frequently use embodied cues for functional self-regulatory purposes (Balcetis and Cole, 2009; Schnall et al., 2010; Bargh and Shalev, 2012; Shalev, 2014). However, using embodied cues as diagnostic inputs (Williams et al., 2009; Ackerman et al., 2010; Meier et al., 2012; Robinson and Fetterman, 2015; Winkielman et al., 2015) may lead to human-robot miscommunications. For example, human speakers expect co-located listeners to link visually perceivable objects and the verbally described references to them. Thus, humans may expect a co-located robot to have the same visual-verbal linking abilities (e.g., look at the green object on the right), thus developers must integrate the robot's visual system with natural language components to enable this flow of communication (Kopp, 2010; Cantrell et al., 2012; Vollmer et al., 2013). Furthermore, there is also anecdotal evidence of human-human communication misunderstandings in complex scenes. For example orientation can be relative to egocentric, or exocentric (absolute or relative) locations. Soldiers for example, are taught to communicate via the exocentric coordinates of the compass rose. However, most humans tend to naturally orient relative to their egocentric perspective, which may be difficult for robots to depict. Interestingly, Cassenti et al. (2012) found that instructors used exocentric references to direct the robot and that it improved their performance relative to egocentric-only commands. To address this communication gap, we argue that shared database, sensors and multiple types of displays and interaction means (e.g., physiological measures, eye tracking, voice, touch, text, button presses etc.) can enrich the robot's capacity of perception and expression. Similarly, to reduce expectation issues, technology can shape the way the user acts on the robot, how individuals understand what to expect from it, and how they can interact with a robot to refine their mutual understanding of the task at hand. Providing the relevant information about the current state of the robot, the progress of the task, and of the surrounding environment, can facilitate successful performance. Similarly, education efforts need to convey the ambiguity of ongoing human-robot communication, particularly the robot's physical and data-driven limitations, and to encourage problem solving and novelty seeking. The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. The research was partially supported by the Helmsley Charitable Trust through the Agricultural, Biological and Cognitive Robotics Center of Ben-Gurion University of the Negev. We thank the editor and reviewers for their helpful comments. Ackerman, J. M., Nocera, C. C., and Bargh, J. A. (2010). Incidental haptic sensations influence social judgments and decisions. Science 328, 1712–1715. doi: 10.1126/science.1189993 PubMed Abstract | CrossRef Full Text | Google Scholar Alami, R., Clodic, A., Montreuil, V., Sisbot, E. A., and Chatila, R. (2005). Task Planning for Human-Robot Interaction. Grenobel: JointsOc-EUSAI Conference. Google Scholar Balcetis, E., and Cole, S. (2009). Body in mind: the role of embodied cognition in self-regulation. Soc. Pers. Psychol. Commun. 759–774. doi: 10.1111/j.1751-9004.2009.00197.x CrossRef Full Text | Google Scholar Bargh, J. A., and Shalev, I. (2012). The substitutability of physical and social warmth in daily life. Emotion 12, 154–162. doi: 10.1037/a0023527 PubMed Abstract | CrossRef Full Text | Google Scholar Baumeister, R. F., and Bargh, J. A. (2014). “Conscious and unconscious: toward an integrative understanding of human life and action,” in Dual Process Theories of the Social Mind, eds J. Sherman, B. Gawronski, and Y. Trope (New York, NY: Guilford Press), 33–49. Boucher, J. D., Pattacini, U., Lelong, A., Bailly, G., Elisei, F., Fagel, S. P. F., et al. (2012). I reach faster when i see you look: gaze effects in human-human and human-robot face-to-face cooperation. Front. Neurorob. 6:3. doi: 10.3389/fnbot.2012.00003 PubMed Abstract | CrossRef Full Text | Google Scholar Cannon-Bowers, J. A., Salas, E., and Converse, S. A. (1993). “Shared mental models in expert team decision making,” in Current Issues in Individual and Group Decision Making, ed N. J. Castellan (Mahwah, NJ: Erlbaum), 221–246. Cantrell, R., Krause, E., Scheutz, M., Zillich, M., and Potapova, E. (2012). “Incremental referent grounding with NLP-biased visual search,” in Proceedings of AAAI 2012 Workshop on Grounding Language for Physical Systems. Carver, C. S., and Scheier, M. F. (1998). On the Self-regulation of Behavior. New York, NY: Cambridge University Press. Google Scholar Cassenti, D. N., Kelley, T. D., Yagoda, R., and Avery, E. (2012). Improvements in robot navigation through operator speech preference. Paladyn 3, 102–111. doi: 10.2478/s13230-012-0100-6 CrossRef Full Text | Google Scholar Chartrand, T. L., and Bargh, J. A. (1999). The chameleon effect: the perception–behavior link and social interaction. J. Pers. Soc. Psychol. 76:893. doi: 10.1037/0022-3514.76.6.893 PubMed Abstract | CrossRef Full Text | Google Scholar Clark, H. H. (2005). Coordinating with each other in a material world. Dis. Stud. 7, 507–525. doi: 10.1177/1461445605054404 CrossRef Full Text | Google Scholar Cohen, E., Ejsmond-Frey, R., Knight, N., and Dunbar, R. I. M. (2010). Rowers' high: behavioural synchrony is correlated with elevated pain thresholds. Biol. Lett. 6, 106–108. doi: 10.1098/rsbl.2009.0670 PubMed Abstract | CrossRef Full Text | Google Scholar Custers, R., and Aarts, H. (2010). The unconscious will: how the pursuit of goals operates outside of conscious awareness. Science 329, 47–50. doi: 10.1126/science.1188595 PubMed Abstract | CrossRef Full Text | Google Scholar Dijksterhuis, A., Preston, J., Wegner, D. M., and Aarts, H. (2008). Effects of subliminal priming of self and God on self-attribution of authorship for events. J. Exp. Soc. Psychol. 44, 2–9. doi: 10.1016/j.jesp.2007.01.003 CrossRef Full Text | Google Scholar Fiore, S. M., Wiltshire, T. J., Lobato, E. J., Jentsch, F. G., Huang, W. H., and Axelrod, B. (2013). Toward understanding social cues and signals in human–robot interaction: effects of robot gaze and proxemic behavior. Front. Psychol. 4:859. doi: 10.3389/fpsyg.2013.00859 PubMed Abstract | CrossRef Full Text | Google Scholar Fishbach, A., and Ferguson, M. F. (2007). “The goal construct in social psychology,” in Social Psychology: Handbook of Basic Principles, eds A. W. Kruglanski and T. E. Higgins (New York, NY: Guilford Press), 490–515. Fishbach, A., and Trope, Y. (2005). The substitutability of external control and self-control in overcoming temptation. J. Exp. Soc. Psychol. 41, 256–270. doi: 10.1016/j.jesp.2004.07.002 CrossRef Full Text | Google Scholar Foster, M. E. (2014). “Validating attention classifiers for multi-party human-robot interaction,” in Proceedings of the 2014 ACM/IEEE International Conference on Human-robot Interaction: Workshop on Attention Models in Robotics. (Bielefeld: ACM Press). Gibson, J. J. (1979). The Ecological Approach to Visual Perception. Boston, MA: Houghton Mifflin. Google Scholar Gollwitzer, P. M., and Bargh, J. A. (eds.). (1996). The Psychology of Action: Linking Cognition and Motivation to Behavior. New York, NY: Guilford Press. Google Scholar Gollwitzer, P. M., and Oettingen, G. (2011). “Planning promotes goal striving,” in Handbook of self-regulation: Research, Theory, and Applications, 2nd edn., eds K. D. Vohs and R. F. Baumeister (New York, NY: Guilford), 162–185. Gray, K., and Wegner, D. M. (2012). Feeling robots and human zombies: mind perception and the uncanny valley. Cognition 125, 125–130. doi: 10.1016/j.cognition.2012.06.007 PubMed Abstract | CrossRef Full Text | Google Scholar Higgins, E. T. (1989). Continuities and discontinuities in self-regulatory and self-evaluative processes: a developmental theory relating self and affect. J. Pers. 57, 407–444. doi: 10.1111/j.1467-6494.1989.tb00488.x PubMed Abstract | CrossRef Full Text | Google Scholar Hoffman, G., Birnbaum, G. E., Reis, H. T., Vanunu, K., and Sass, O. (2014). “Robot responsiveness to human disclosure affects social impression and appeal,” in Proceedings of the 9th ACM/IEEE International Conference on Human-Robot Interaction. Klimoski, R., and Mohammed, S. (1994). Team mental model: construct or metaphor? J. Man. 20, 403–437. Google Scholar Koole, S. L., and Veenstra, L. (2015). Does emotion regulation occur only inside people's heads? Towards a situated cognition analysis of emotion-regulatory dynamics. Psychol. Inq. 26, 61–68. doi: 10.1080/1047840X.2015.964657 CrossRef Full Text | Google Scholar Kopp, S. (2010). Social resonance and embodied coordination in face-to-face conversation with artificial interlocutors Speech Commun. 52, 587–597. doi: 10.1016/j.specom.2010.02.007 CrossRef Full Text | Google Scholar Kozlowski, S. W. J., and Bell, B. S. (2003). “Work groups and teams in organizations,” in Handbook of Psychology: Industrial and Organizational Psychology eds W. C. Borman, D. R. Ilgen, R. J. Klimoski (London: Wiley), 333–375. Kruglanski, A. W., Shah, J. Y., Fishbach, A., Friedman, R., Chun, W. Y., and Sleeth-Keppler, D. (2002). A theory of goal systems. Adv. Exp. Soc. Psychol. 34, 331–378. doi: 10.1016/s0065-2601(02)80008-9 CrossRef Full Text | Google Scholar Leander, N. P., Chartrand, T. L., and Wood, W. (2010). Mind your mannerisms: behavioral mimicry elicits stereotype conformity. J. Exp. Soc. Psychol. 47, 195–201. doi: 10.1016/j.jesp.2010.09.002 CrossRef Full Text | Google Scholar Loersch, C., and Payne, B. K. (2011). The situated inference model an integrative account of the effects of primes on perception, behavior, and motivation. Pers. Psychol. Sci. 6, 234–252. doi: 10.1177/1745691611406921 CrossRef Full Text | Google Scholar Lohan, K. S., Griffiths, S. S., Sciutti, A., Partmann, T. C., and Rohlfing, K. J. (2014). Co−development of manner and path concepts in language, action, and eye−gaze behavior. Top. Cogn. Sci. 6, 492–512. doi: 10.1111/tops.12098 PubMed Abstract | CrossRef Full Text | Google Scholar Loth, S., Huth, K., and de Ruiter, J. P. (2013). Automatic detection of service initiation signals used in bars. Front. Psychol. 4:557. doi: 10.3389/fpsyg.2013.00557 PubMed Abstract | CrossRef Full Text | Google Scholar Meier, B. P., Schnall, S., Schwarz, N., and Bargh, J. A. (2012). Embodiment in social psychology. Top. Cogn. Sci. 4, 705–716. doi: 10.1111/j.1756-8765.2012.01212.x PubMed Abstract | CrossRef Full Text | Google Scholar Paladino, M. P., Mazzurega, M., Pavani, F., and Schubert, T. W. (2010). Synchronous multisensory stimulation blurs self-other boundaries. Psychol. Sci. 21, 1202–1207. doi: 10.1177/0956797610379234 PubMed Abstract | CrossRef Full Text | Google Scholar Pentland, A. A. (2007). Social signal processing. IEEE Signal Process. Mag. 24, 108–111. doi: 10.1109/MSP.2007.4286569 CrossRef Full Text | Google Scholar Robinson, M. D., and Fetterman, A. K. (2015). The embodiment of success and failure as forward versus backward movements. PLoS ONE 10:e0117285. doi: 10.1371/journal.pone.0117285 PubMed Abstract | CrossRef Full Text | Google Scholar Salas, E., Cooke, N. J., and Gorman, J. C. (2010). The science of team performance: progress and the need for more. Int. J. Hum. Comput. Stud. 45, 75–104. doi: 10.1177/0018720810374614 CrossRef Full Text | Google Scholar Salas, E., Dickinson, T. L., Converse, S. A., and Tannenbaum, S. I. (1992). “Toward an understanding of team performance and training,” in Teams: Their Training and Performance, eds R. J. Swezey and E. Salas (Norwood, NJ: Ablex), 3–29. Google Scholar Schnall, S., Zadra, J. R., and Proffitt, D. R. (2010). Direct evidence for the economy of action: glucose and the perception of geographical slant. Perception 39, 464. doi: 10.1068/p6445 PubMed Abstract | CrossRef Full Text | Google Scholar Sebanz, N., and Knoblich, G. (2009). Prediction in joint action: what, when, and where. Top. Cogn. Sci. 1, 353–367. doi: 10.1111/j.1756-8765.2009.01024.x PubMed Abstract | CrossRef Full Text | Google Scholar Shah, J. Y., Friedman, R., and Kruglanski, A. W. (2002). Forgetting all else: on the antecedents and consequences of goal shielding. J. Pers. Soc. Psychol. 83:1261. doi: 10.1037/0022-3514.83.6.1261 PubMed Abstract | CrossRef Full Text | Google Scholar Shalev, I. (2014). Implicit energy loss: embodied dryness cues influence vitality and depletion. J. Cons. Psychol. 24, 260–270. doi: 10.1016/j.jcps.2013.09.011 CrossRef Full Text | Google Scholar Shalev, I. (2015). The architecture of embodied cue integration: insight from the “motivation as cognition” perspective. Front. Psychol. 6:658. doi: 10.3389/fpsyg.2015.00658 PubMed Abstract | CrossRef Full Text | Google Scholar Talmy, L. (1991). “Path to realization: a typology of event conflation,” in Proceedings of the 17th Annual Meeting of the Berkeley Linguistics Society, eds L. A. Sutton, C. Johnson and R. Shields (Berkeley, CA: Berkeley Linguistic Society), 480–519. Vallacher, R. R., and Wegner, D. M. (1987). What do people think they're doing? Action identification and human behavior. Psychol. Rev. 94:3. doi: 10.1037/0033-295X.94.1.3 CrossRef Full Text | Google Scholar Vallacher, R. R., and Wegner, D. M. (1989). Levels of personal agency: individual variation in action identification. J. Per. Soc. Psychol. 57:660. doi: 10.1037/0022-3514.57.4.660 PubMed Abstract | CrossRef Full Text | Google Scholar van Baaren, R. B., Maddux, W. W., Chartrand, T. L., DeBouter, C., and van Knippenberg, A. (2003). It takes two to mimic: behavioral consequences of self-construals. J. Pers. Soc. Psychol. 84, 1093–1102. doi: 10.1037/0022-3514.84.5.1093 PubMed Abstract | CrossRef Full Text | Google Scholar Vinciarelli, M., Pantic, D., Heylen, C., Pelachaud, I., Poggi, F., D'ericco, M., et al. (2012). Bridging the gap between social animal and unsocial machine: a survey of social signal processing. IEEE Trans. Affect. Comput. 3, 69–87. doi: 10.1109/T-AFFC.2011.27 CrossRef Full Text | Google Scholar Vollmer, A.-L., Wrede, B., Rohlfing, K. J., and Cangelosi, A. (2013). Do beliefs about a robot's capabilities influence alignment to its actions? Proc. EpiRob/ICDL 2013, 1–6. doi: 10.1109/DevLrn.2013.6652521 CrossRef Full Text | Google Scholar Williams, L. E., Huang, J. Y., and Bargh, J. A. (2009). The scaffolded mind: higher mental processes are grounded in early experience of the physical world. Eur. J. Soc. Psychol. 39, 1257–1267. doi: 10.1002/ejsp.665 PubMed Abstract | CrossRef Full Text | Google Scholar Winkielman, P., Niedenthal, P., Wielgosz, J., Eelen, J., and Kavanagh, L. C. (2015). “Embodiment of cognition and emotion,” in APA handbooks in psychology. APA handbook of personality and social psychology, Vol. 1. Attitudes and social cognition eds M. Mikulincer, P. R. Shaver, E. Borgida and J. A. Bargh (Washington, DC: American Psychological Association), 151–175. doi: 10.1037/14341-004 CrossRef Full Text Keywords: human-robot interaction (HRI), embodied cognition, self-regulation, goal pursuit, teamwork Citation: Shalev I and Oron-Gilad T (2015) What do we think we are doing: principles of coupled self-regulation in human-robot interaction (HRI). Front. Psychol. 6:929. doi: 10.3389/fpsyg.2015.00929 Received: 15 March 2015; Accepted: 22 June 2015; Published: 08 July 2015. Edited by: Reviewed by: Copyright © 2015 Shalev and Oron-Gilad. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. *Correspondence: Idit Shalev, [email protected]
Tim D. Smithies, , Eoin Conroy, , Magdalena Kowal, Mark J. Campbell
Published: 5 May 2020
Frontiers in Psychology, Volume 11; https://doi.org/10.3389/fpsyg.2020.00883

Abstract:
Esports has experienced unprecedented growth recently and with it, the proposition for aspiring gamers to pursue a professional esport career has become increasingly attractive. “Esports” are video-games played competitively (and often professionally) through the means of cyberspace (Campbell et al., 2018) and are an important fixture in the overall gaming industry, which is estimated to be worth more than 120 billion US$ (Takahashi, 2020). The exponential rise in popularity has led to the inclusion of two esports (Rocket League and Street Fighter V) in an International Olympics Council sanctioned tournament before the 2020 Tokyo Olympic games (Martinello, 2019). Despite the appeal of esports as a profession, aspiring esport athletes face many obstacles that can threaten their prospective career timespan, and present post-career difficulties. To date, very limited formal exploration exists into this challenge; thus, this grand field challenge aims to explore the difficulties faced by esport athletes. It also highlights the unique skillsets and experience acquired during a professional esports career, and the value these could offer to alternate high performance professions. Like all occupations, an esports career depends on financial and job security. Although some professional athletes in tier-1 (the highest level of competition) leagues within popular esports enjoy financial stability from yearly contracts (Esports Mention, 2019), athletes in less popular esports and aspiring gamers not yet competing in tier-1 leagues are not afforded this luxury. Additionally, outside tier-1 tournaments, prize money distribution is such that winners are more greatly rewarded at the expense of other participants (Coates and Parshakov, 2016). Moreover, protections are limited for esport athletes as they have not yet been able to unionize. This makes job security remarkably fragile, particularly given the high athlete replaceability, with extreme cases of top esport athletes being dropped from teams at post-victory celebrations (Van Allen, 2018). To further compound these challenges, the average career of typical esports athletes' is remarkably short; with about one-in-five professional esport athletes' careers lasting 2 years or longer (Ward and Harmon, 2019). This short career length is largely due to the difficulty of becoming and remaining a top esport athlete, particularly given the volatility of team rosters. Esports performance is reliant on the ability to rapidly and accurately respond to complex visual stimuli, which begins to decline past 24 years of age (Thompson et al., 2014). As such, one's timeframe for peak esports performance is limited. The time commitment and rigor required for elite esports performance has resulted in many cases of burnout and injury, causing early retirement (Salo, 2017). Additionally, adolescent esport athletes often sacrifice educational opportunities to pursue their careers (Hollist, 2015), hampering their ability to pursue alternate careers post-retirement. In summary, there appears to be a narrow timeframe for financial success for esports athletes, who may jeopardize their post-retirement opportunities to take that window. Despite the current pitfalls of an esports profession, esports athletes possess a unique range of specialized skills and experiences that we argue are highly sought after in many contemporary professions. Such attributes include digital intelligence, experience and expertise in prolonged human computer interaction performed in a seated posture, skillful and efficient communication, and perhaps most notably, enhanced cognitive abilities. In this section, we outline some of these traits that this population possess. Fundamentally, esports involve human-computer interactions with an adapting computer program to produce outcome-defining events within a virtual gameplay environment (Hamari and Sjöblom, 2017). Higher-level esport athletes perform faster and with more complexity than their less-skilled counterparts (Avontuur et al., 2013; Buckley et al., 2017). Esports are a type of high-performance computing that requires “digital intelligence” to provide a competitive advantage, such as knowledge of, and proficiency with, hardware components (Claypool and Claypool, 2007). Additionally, esports athletes undertake long continuous bouts (often >3 h) fixating on computer monitors in a seated posture during training and competition (DiFrancisco-Donoghue et al., 2019). It is well-established that this prolonged sitting can result in lower back discomfort and impaired vascular function (Dunk and Callaghan, 2010; Credeur et al., 2019). Moreover, frequent computer monitor use can lead to “computer vision syndrome,” associated with temporary eye discomfort (Blehm et al., 2005). Although little work has investigated these physiological effects in esports athletes, it may be that esports athletes have developed strategies to maintain performance despite these issues. Esport teams are a unique hybrid of a high-performance action team engaging in computed supported cooperative work (CSCW), a combination not regularly seen in more traditional team environments (Freeman and Wohn, 2019). Given that most esports are team-based, effective team cohesion and communication are essential for success. Communication within elite esports is overwhelmingly verbal (Lipovaya et al., 2018). To maximize efficiency and effectiveness of this communication, athletes must be proficient in utilizing rigid phraseologies (Lipovaya et al., 2018; Freeman and Wohn, 2019). Team strategies and individual roles must be effectively communicated prior-to and during competition to ensure successful performance (Lipovaya et al., 2018; Freeman and Wohn, 2019). First Person Shooter (FPS) and Multiplayer Online Battle Arena (MOBA) games, which collectively comprise the majority of major esports and are known as action video-games (AVGs), are cognitively demanding (Campbell et al., 2018). The demand that esports places on athletes has resulted in a growing body of research demonstrating that gamers possess enhanced cognitive abilities compared to non-gamers (Kowal et al., 2018; Large et al., 2019); due to this, esport athletes have been referred to as “cognitive athletes” (Campbell et al., 2018). Existing literature indicates that experienced AVG players (AVGPs) have enhanced spatial and temporal visual perception (Green and Bavelier, 2007; West et al., 2008; Li et al., 2009, 2010; Appelbaum et al., 2013). Further, AVGPs possess greater attentional resources, facilitating performance improvements on tasks with large attentional-demands (Bavelier et al., 2012a; Krishnan et al., 2013). Additionally, AVGPs can also better control their attentional resources (Chisholm et al., 2010; Mishra et al., 2011; Bavelier et al., 2012a; Chisholm and Kingstone, 2012; Green et al., 2012; Krishnan et al., 2013; Cain et al., 2014; Föcker et al., 2018) and allocate them over a wider visual field-of-view (Dale et al., 2019). Lastly, AVGPs demonstrate high capacity to integrate visual and auditory information (Donohue et al., 2010). In addition to enhanced perception and attentional capabilities, AVGPs have been demonstrated to have faster overall response times across a diverse range of tasks, which is believed to represent general enhancements in cognitive throughput (Castel et al., 2005; West et al., 2008; Dye et al., 2009; Hubert-Wallander et al., 2011; Bavelier et al., 2012a; Green et al., 2012; Wu and Spence, 2013; Föcker et al., 2018). It has been suggested that these improved perceptual, attentional, and processing speed abilities provide gamers an enhanced capacity to learn as well (Bavelier et al., 2012b). Although their careers are tenuous and short-lived, the inherent skill-sets possessed by esports athletes are highly desirable in multiple contemporary professions. To demonstrate this, we queried the Occupational Information Network to determine professions sharing expertise and experience with esport athletes (O*net, queried 12/12/19). O*net is a free database of occupation-specific descriptors, developed with the sponsorship of the U.S. Department of Labor/Employment and Training Administration (USDOL/ETA), to help jobseekers match careers to their skill-sets. Within the database, almost 1,000 professions are ranked according to defined categories of abilities/experience; we queried those categories related to the expertise of esport athletes. Following our searches, two professions very regularly (over 40%) appeared (top 30 most relevant occupations; see Supplementary Table 1 for search details): aircraft pilots (hereafter simply referred to as “pilots”) and air traffic controllers (ATCs). Additionally, previous research demonstrating comparable performance of AVGPs to military combat pilots on simulated Unmanned Aerial System (UAS) operations (McKinley et al., 2011) led us to include this profession. The following sections explore how the skill-sets and experience required for success in esports could be highly valued in these professions (see Figure 1). Figure 1. Prominent skills and experiences possessed by esport athletes (left) and a Venn diagram (right) visually demonstrating the mutual desire for esport-related skill and experience by military UAS (drone) operators, air traffic controllers, and pilots. Expert interfacing with complex computerized systems is essential for the outlined professions, with ATCs and UAS operators, in a similar manner to esports, using computer monitors as primary outputs. For ATCs, the quality of such human-computer interface interactions is integral to overall performance (Chang and Yeh, 2010). Furthermore, McKinley et al. (2011) noted that gamers were highly proficient at using “game-like” UAS interfaces, supporting the performance benefits of interface familiarity. All of the outlined professions are also performed while seated, with pilots specifically remaining in such a posture for several hours at a time (Lusted et al., 1994). While pilots do not fixate on screens for prolonged periods, ATCs and UAS operators do, and may experience “computer vision syndrome” as a result. Given that esport athletes often perform during long bouts of sitting and screen fixation, they may be better suited to maintaining high task performance in jobs that appear to also have these challenges. Pilots and ATCs must work synergistically to maximize safety in operations where communication, much like esports, is primarily digital (CSCW). Very specific phraseology is used in aviation to optimize communication efficiency (Campbell-Laird, 2004). Such consistency is vital, as coordination and communication errors are the leading cause of air traffic accidents (Isaac and Ruitenberg, 2017). Similarly, UAS operators are required to have frequent and concise verbal communication with ATCs, ground units, and other aircraft (McKinley et al., 2011). The key similarities between esports athletes, pilots, ATCs and UAS operators lie in their cognitive abilities. Rapid identification (perception and attention) and processing (cognitive throughput) of information is vital for safety and operational performance. For ATCs and pilots, timely recognition and response to warnings on one of many displays (often in visual peripheries) can prevent hundreds of casualties, while timely localization and action toward a target can define operational success or failure among UAS operators. Important information for all three professions is invariably concealed among an array or “clutter” of stimuli, rendering attentional control as critical. Moreover, individuals in these professions are often required to simultaneously attend to multiple stimuli at once (multitasking); particularly ATCs, who constantly must manage airspaces containing numerous aircraft. Information may be visual (displays), haptic and auditory (alarms, warnings, and verbal), placing importance on multisensory integration. Given such intense demands, attentional errors are among the most common errors for ATCs (Pape et al., 2001), and cognitive processing/decision making errors constitute most “pilot error” accidents (Adams and Ericsson, 2000); both of which can result in numerous casualties. To mitigate such risks, pre-training assessments for these professions regularly include assessing these aforementioned cognitive attributes; given the cognitive proficiency of esport athletes, they may perform well on such tests, demonstrating high suitability for these professions. It must be acknowledged that the skill-sets/experiences possessed by esport athletes are not exclusively beneficial for pilots, ATCs, and UAS operators, and may be favorable for any occupations which share similar workplace demands, communication, and cognitive requirements. The rise of esports has resulted in the emergence of a population of uniquely skilled young individuals. These “cognitive athletes” can quickly perceive and process large amounts of information, while simultaneously demonstrating better attentional control. Moreover, they are strong communicators and work well in team environments, particularly through digital means and in high-performance computing contexts. Lastly, they are notably proficient with human-computer interfaces, and are experienced with working in a seated posture for extended periods. Unfortunately, given the nature of esports as a profession, most esport athletes experience a short, financially unstable career, with limited post-retirement opportunities. Here, we have highlighted the shared importance of the unique skills and experiences possessed by esport athletes and how they may be preferentially valued for three exemplar professions; pilots, ATCs, and military UAS operators. High-performance in these three professions is critical, as errors can pose large financial and human costs. Overall, this work poses a challenge to the esports, scientific and industrial communities, to demonstrate how best to leverage the unique abilities of esports athletes to facilitate their life after esports and add value to professions seeking individuals with these unique skillsets. Doing so could result in more suitable personnel occupying the abovementioned industries and would be highly beneficial to the distinctly vulnerable population of esport athletes. All authors contributed to the conception of the field challenge. TS wrote the manuscript. All authors contributed to manuscript revision, read, and approved the submitted version. This work was supported with the financial support of the Science Foundation Ireland grant 13/RC/2094 and co-funded under the European Regional Development Fund through the Southern & Eastern Regional Operational Programme to Lero—the SFI Centre for Software Research (www.lero.ie). TS was receiving funding from the Irish Research Council Employment-Based Postgraduate Program Scholarship (EBPPG/2019/21), with Logitech as the Employment Partner. The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpsyg.2020.00883/full#supplementary-material Adams, R. J., and Ericsson, A. E. (2000). Introduction to cognitive processes of expert pilots. Hum. Perf. Extrem. Environ. 5, 44–62. doi: 10.7771/2327-2937.1006 PubMed Abstract | CrossRef Full Text | Google Scholar Appelbaum, L. G., Cain, M. S., Darling, E. F., and Mitroff, S. R. (2013). Action video game playing is associated with improved visual sensitivity, but not alterations in visual sensory memory. Atten. Percept. Psychophys. 75, 1161–1167. doi: 10.3758/s13414-013-0472-7 CrossRef Full Text | Google Scholar Avontuur, T., Spronck, P., and Van Zaanen, M. (2013). “Player skill modeling in Starcraft II,” in Ninth Artificial Intelligence and Interactive Digital Entertainment Conference, eds G. Sukthankar and I. Horswil (Boston, MA: AAAI Press), 2–8. Google Scholar Bavelier, D., Achtman, R. L., Mani, M., and Föcker, J. (2012a). Neural bases of selective attention in action video game players. Vision Res. 61, 132–143. doi: 10.1016/j.visres.2011.08.007 PubMed Abstract | CrossRef Full Text | Google Scholar Bavelier, D., Shawn Green, C., Pouget, A., and Schrater, P. (2012b). Brain plasticity through the life span: learning to learn and action video games. Annu. Rev. Neurosci. 35, 391–416. doi: 10.1146/annurev-neuro-060909-152832 PubMed Abstract | CrossRef Full Text | Google Scholar Blehm, C., Vishnu, S., Khattak, A., Mitra, S., and Yee, R. W. (2005). Computer vision syndrome: a review. Surv. Ophthalmol. 50, 253–262. doi: 10.1016/j.survophthal.2005.02.008 PubMed Abstract | CrossRef Full Text | Google Scholar Buckley, D., Chen, K., and Knowles, J. (2017). Rapid skill capture in a first-person shooter. IEEE Trans. Comput. Intell. AI Games 9, 63–75. doi: 10.1109/TCIAIG.2015.2494849 CrossRef Full Text | Google Scholar Cain, M., Prinzmetal, W., Shimamura, A., and Landau, A. (2014). Improved control of exogenous attention in action video game players. Front. Psychol. 5:69. doi: 10.3389/fpsyg.2014.00069 PubMed Abstract | CrossRef Full Text | Google Scholar Campbell, M. J., Toth, A. J., Moran, A. P., Kowal, M., and Exton, C. (2018). eSports: a new window on neurocognitive expertise? Prog. Brain Res. 240, 161–174. doi: 10.1016/bs.pbr.2018.09.006 PubMed Abstract | CrossRef Full Text | Google Scholar Campbell-Laird, K. (2004). “Aviation English: a review of the language of International Civil Aviation,” in International Professional Communication Conference (Minneapolis, MN: IPCC), 253–261. Google Scholar Castel, A. D., Pratt, J., and Drummond, E. (2005). The effects of action video game experience on the time course of inhibition of return and the efficiency of visual search. Acta Psychol. 119, 217–230. doi: 10.1016/j.actpsy.2005.02.004 PubMed Abstract | CrossRef Full Text | Google Scholar Chang, Y. H., and Yeh, C. H. (2010). Human performance interfaces in air traffic control. Appl. Ergon. 41, 123–129. doi: 10.1016/j.apergo.2009.06.002 PubMed Abstract | CrossRef Full Text | Google Scholar Chisholm, J. D., Hickey, C., Theeuwes, J., and Kingstone, A. (2010). Reduced attentional capture in action video game players. Atten. Percept. Psychophys. 72, 667–671. doi: 10.3758/APP.72.3.667 PubMed Abstract | CrossRef Full Text | Google Scholar Chisholm, J. D., and Kingstone, A. (2012). Improved top-down control reduces oculomotor capture: the case of action video game players. Atten. Percept. Psychophys. 74, 257–262. doi: 10.3758/s13414-011-0253-0 PubMed Abstract | CrossRef Full Text | Google Scholar Claypool, K. T., and Claypool, M. (2007). On frame rate and player performance in first person shooter games. Multimedia Syst. 13, 3–17. doi: 10.1007/s00530-007-0081-1 CrossRef Full Text | Google Scholar Coates, D., and Parshakov, P. (2016). “Team vs. individual tournaments: evidence from prize structure in eSports,” in Higher School of Economics Research Paper No. WP BRP 138. doi: 10.2139/ssrn.2787819 CrossRef Full Text | Google Scholar Credeur, D. P., Miller, S. M., Jones, R., Stoner, L., Dolbow, D. R., Fryer, S. M., et al. (2019). Impact of prolonged sitting on peripheral and central vascular health. Am. J. Cardiol. 123, 260–266. doi: 10.1016/j.amjcard.2018.10.014 PubMed Abstract | CrossRef Full Text | Google Scholar Dale, G., Kattner, F., Bavelier, D., and Green, C. S. (2019). Cognitive abilities of action video game and role-playing video game players: data from a massive open online course. Psychol. Pop. Media Cult. doi: 10.1037/ppm0000237. [Epub ahead of print]. CrossRef Full Text | Google Scholar DiFrancisco-Donoghue, J., Balentine, J., Schmidt, G., and Zwibel, H. (2019). Managing the health of the eSport athlete: an integrated health management model. BMJ Open Sport Exerc. Med. 5:e000467. doi: 10.1136/bmjsem-2018-000467 PubMed Abstract | CrossRef Full Text | Google Scholar Donohue, S. E., Woldorff, M. G., and Mitroff, S. R. (2010). Video game players show more precise multisensory temporal processing abilities. Atten. Percept. Psychophys. 72, 1120–1129. doi: 10.3758/APP.72.4.1120 PubMed Abstract | CrossRef Full Text | Google Scholar Dunk, N. M., and Callaghan, J. P. (2010). Lumbar spine movement patterns during prolonged sitting differentiate low back pain developers from matched asymptomatic controls. Work 35, 3–14. doi: 10.3233/WOR-2010-0953 PubMed Abstract | CrossRef Full Text | Google Scholar Dye, M. W. G., Green, C. S., and Bavelier, D. (2009). The development of attention skills in action video game players. Neuropsychologia 47, 1780–1789. doi: 10.1016/j.neuropsychologia.2009.02.002 PubMed Abstract | CrossRef Full Text | Google Scholar Esports Mention (2019). How Much Do Esports Players Make? Esports Mention. Available online at: https://esportsmention.com/esports/insights/how-much-do-esports-players-make/ (accessed January 14, 2020). Google Scholar Föcker, J., Mortazavi, M., Khoe, W., Hillyard, S. A., and Bavelier, D. (2018). Neural correlates of enhanced visual attentional control in action video game players: an event-related potential study. J. Cogn. Neurosci. 31, 377–389. doi: 10.1162/jocn_a_01230 PubMed Abstract | CrossRef Full Text | Google Scholar Freeman, G., and Wohn, D. Y. (2019). Understanding eSports team formation and coordination. CSCW 28, 95–126. doi: 10.1007/s10606-017-9299-4 CrossRef Full Text | Google Scholar Green, C. S., and Bavelier, D. (2007). Action-video-game experience alters the spatial resolution of vision: research article. Psychol. Sci. 18, 88–94. doi: 10.1111/j.1467-9280.2007.01853.x CrossRef Full Text | Google Scholar Green, C. S., Sugarman, M. A., Medford, K., Klobusicky, E., and Bavelier, D. (2012). The effect of action video game experience on task-switching. Comput. Hum. Behav. 28, 984–994. doi: 10.1016/j.chb.2011.12.020 PubMed Abstract | CrossRef Full Text | Google Scholar Hamari, J., and Sjöblom, M. (2017). What is eSports and why do people watch it? Intern. Res. 27, 211–232. doi: 10.1108/IntR-04-2016-0085 CrossRef Full Text | Google Scholar Hollist, K. E. (2015). Time to be grown-ups about video gaming: the rising eSports industry and the need for regulation. Arizona Law Rev. 57:823. Google Scholar Hubert-Wallander, B., Green, C. S., Sugarman, M., and Bavelier, D. (2011). Changes in search rate but not in the dynamics of exogenous attention in action videogame players. Atten. Percept. Psychophys. 73, 2399–2412. doi: 10.3758/s13414-011-0194-7 CrossRef Full Text | Google Scholar Isaac, A. R., and Ruitenberg, B. (2017). Air Traffic Control: Human Performance Factors. London: Routledge. Google Scholar Kowal, M., Toth, A. J., Exton, C., and Campbell, M. J. (2018). Different cognitive abilities displayed by action video gamers and non-gamers. Comput. Hum. Behav. 88, 255–262. doi: 10.1016/j.chb.2018.07.010 CrossRef Full Text | Google Scholar Krishnan, L., Kang, A., Sperling, G., and Srinivasan, R. (2013). Neural strategies for selective attention distinguish fast-action video game players. Brain Topogr. 26, 83–97. doi: 10.1007/s10548-012-0232-3 PubMed Abstract | CrossRef Full Text | Google Scholar Large, A. M., Bediou, B., Cekic, S., Hart, Y., Bavelier, D., and Green, C. S. (2019). Cognitive and behavioral correlates of achievement in a complex multi-player video game. Media Commun. 7, 198–212. doi: 10.17645/mac.v7i4.2314 CrossRef Full Text | Google Scholar Li, R., Polat, U., Makous, W., and Bavelier, D. (2009). Enhancing the contrast sensitivity function through action video game training. Nat. Neurosci. 12, 549–551. doi: 10.1038/nn.2296 PubMed Abstract | CrossRef Full Text | Google Scholar Li, R., Polat, U., Scalzo, F., and Bavelier, D. (2010). Reducing backward masking through action game training. J. Vision 10:333. doi: 10.1167/10.14.33 PubMed Abstract | CrossRef Full Text | Google Scholar Lipovaya, V., Lima, Y., Grillo, P., Barbosa, C. E., de Souza, J. M., Duarte, F. J., et al. (2018). “Coordination, communication, and competition in eSports: a comparative analysis of teams in two action games,” in Proceedings of 16th European Conference on Computer-Supported Cooperative Work-Exploratory Papers (Nancy: European Society for Socially Embedded Technologies (EUSSET)]. Google Scholar Lusted, M., Healey, S., and Mandryk, J. A. (1994). Evaluation of the seating of Qantas flight deck crew. Appl. Ergon. 25, 275–282. doi: 10.1016/0003-6870(94)90041-8 PubMed Abstract | CrossRef Full Text | Google Scholar Martinello, E. (2019). Intel World Open brings esports to 2020 Summer Olympics. Esports Insider. Available online at: https://esportsinsider.com/2019/09/intel-world-open-olympics/ (accessed January 14, 2020). Google Scholar McKinley, R. A., McIntire, L. K., and Funke, M. A. (2011). Operator selection for unmanned aerial systems: comparing video game players and pilots. Aviat. Space Environ. Med. 82, 635–642. doi: 10.3357/ASEM.2958.2011 CrossRef Full Text | Google Scholar Mishra, J., Zinni, M., Bavelier, D., and Hillyard, S. A. (2011). Neural basis of superior performance of action videogame players in an attention-demanding task. J. Neurosci. 31, 992–998. doi: 10.1523/JNEUROSCI.4834-10.2011 PubMed Abstract | CrossRef Full Text | Google Scholar Pape, A. M., Wiegmann, D. A., and Shappell, S. A. (2001). “Air traffic control (ATC) related accidents and incidents: a human factors analysis,” in 11th International Symposium on Aviation Psychology (Columbus, OH: The Ohio State University). Google Scholar Salo, M. (2017). Career transitions of eSports Athletes: a proposal for a research framework. IJGCMS 9, 22–32. doi: 10.4018/IJGCMS.2017040102 CrossRef Full Text | Google Scholar Takahashi, D. (2020). SuperData: Games Hit $120.1 Billion in 2019, With Fortnite Topping $1.8 Billion. Games: Venture Beat. Available online at: https://venturebeat.com/2020/01/02/superdata-games-hit-120-1-billion-in-2019-with-fortnite-topping-1-8-billion/ (accessed January 24, 2020). Google Scholar Thompson, J. J., Blair, M. R., and Henrey, A. J. (2014). Over the hill at 24: persistent age-related cognitive-motor decline in reaction times in an ecologically valid video game task begins in early adulthood. PLoS ONE 9:e94215. doi: 10.1371/journal.pone.0094215 PubMed Abstract | CrossRef Full Text | Google Scholar Van Allen, E. (2018). Report: Dota Team Told Player He Was Cut During Pizza Party. Kotaku. Available online at: https://compete.kotaku.com/report-dota-team-told-player-he-was-cut-during-pizza-p-1822249299 (accessed December 20, 2019). Google Scholar Ward, M. R., and Harmon, A. D. (2019). ESport superstars. J. Sports Econ. 20, 987–1013. doi: 10.1177/1527002519859417 CrossRef Full Text | Google Scholar West, G. L., Stevens, S. A., Pun, C., and Pratt, J. (2008). Visuospatial experience modulates attentional capture: evidence from action video game players. J. Vision 8, 13.11–19. doi: 10.1167/8.16.13 PubMed Abstract | CrossRef Full Text | Google Scholar Wu, S., and Spence, I. (2013). Playing shooter and driving videogames improves top-down guidance in visual search. Attent. Percept. Psychophys. 75, 673–686. doi: 10.3758/s13414-013-0440-2 PubMed Abstract | CrossRef Full Text | Google Scholar Keywords: video-games, career transitions, esport retirement, cognitive abilities, digital abilities Citation: Smithies TD, Toth AJ, Conroy E, Ramsbottom N, Kowal M and Campbell MJ (2020) Life After Esports: A Grand Field Challenge. Front. Psychol. 11:883. doi: 10.3389/fpsyg.2020.00883 Received: 12 February 2020; Accepted: 09 April 2020; Published: 05 May 2020. Edited by: Reviewed by: Copyright © 2020 Smithies, Toth, Conroy, Ramsbottom, Kowal and Campbell. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. *Correspondence: Niall Ramsbottom, [email protected] ORCID: Tim D. Smithies orcid.org/0000-0002-8026-5134 Adam J. Toth orcid.org/0000-0003-2193-0138 Eoin Conroy orcid.org/0000-0001-8653-1272 Niall Ramsbottom orcid.org/0000-0002-2992-6136 Magdalena Kowal orcid.org/0000-0002-4768-4900 Mark J. Campbell orcid.org/0000-0001-9607-7675
, Lei Deng, Yansong Chua, Peng Li, Emre O. Neftci, Haizhou Li
Published: 15 April 2020
Frontiers in Neuroscience, Volume 14; https://doi.org/10.3389/fnins.2020.00276

Abstract:
Editorial on the Research TopicSpiking Neural Network Learning, Benchmarking, Programming and Executing A spiking neural network (SNN), a type of brain-inspired neural network, mimics the biological brain, specifically, its neural codes, neuro-dynamics, and circuitry. SNNs have garnered great interest in both Artificial Intelligence (AI) and neuroscience communities given its great potential in biologically realistic modeling of human cognition and development of energy efficient, event-driven machine learning hardware (Pei et al., 2019; Roy et al., 2019). Significant progress has been made across a wide spectrum of AI fields, such as image processing, speech recognition, and machine translation. They are largely driven by the advance in Artificial Neural Networks (ANN) in systematic learning theories, explicit benchmarks with various tasks and data sets, friendly programming tools [e.g., TensorFlow (Abadi et al., 2016) and Pytorch (Paszke et al., 2019) machine learning tools], and efficient processing platforms [e.g., graphics processing unit (GPU) and tensor processing unit (TPU) (Jouppi et al., 2017)]. In comparison, SNNs are still at an early stage in these aspects. To further exploit the advantages of SNNs and attract more researchers to contribute in this field, we proposed a Research Topic in Frontiers in Neuroscience to discuss the main challenges and future prospects of SNNs, emphasizing on its “Learning algorithms, Benchmarking, Programming, and Executing.” We are confident that SNNs will play a critical role in the development of energy efficient machine learning devices through algorithm-hardware co-design. This Research Topic brings together researchers of different disciplines in order to present their recent work in SNNs. We received 22 submissions worldwide and accepted 15 papers. The scope of the accepted papers covers learning algorithms, model efficiency, programming tools, and neuromorphic hardware. Learning algorithms play perhaps the most important role in AI techniques. Machine learning algorithms, in particular those for deep neural networks (DNN), have become the standard bearer in a wide spectrum of AI tasks. Some of the more common learning algorithms include backpropagation (Hecht-Nielsen, 1992), stochastic gradient descent (SGD) (Bottou, 2012), and ADAM optimization (Kingma and Ba, 2014). Other techniques such as batch normalization (Ioffe and Szegedy, 2015) and distributed training (Dean et al., 2012) facilitate learning in DNNs and enable them to be applied in various real-world applications. In comparison, there are relatively fewer SNN learning algorithms and techniques. Existing SNN learning algorithms fall into three categories: unsupervised learning algorithms such as the original spike timing dependent plasticity (STDP) (Querlioz et al., 2013; Diehl and Cook, 2015; Kheradpisheh et al., 2016), indirect supervised learning such as ANN-to-SNN conversion (O'Connor et al., 2013; Pérez-Carrasco et al., 2013; Diehl et al., 2015; Sengupta et al., 2019), and direct supervised learning such as spatiotemporal backpropagation (Wu et al., 2018, 2019a,b). We note that progress in STDP research includes introducing a reward or supervision signal such as spike timing which, in combination with this third factor, dictates the weight changes (Paugam-Moisy et al., 2006; Franosch et al., 2013). Despite the progress made, no algorithm can yet train a very deep SNN efficiently, which has become almost the holy grail of our field. Below, we briefly summarize the accepted algorithm papers in this Research Topic. Inspired by the mammalian olfactory system, Borthakur and Cleland develop an SNN model trained using STDP for signal restoration and identification. It is broadly applicable to sensor array inputs. Luo et al. propose a new weight update mechanism that adjusts synaptic weights, leading to the first wrong output spike-timing to classify input spike trains with time-sensitive information accurately. He et al. divide the learning (weight training) process into two phases: the structure formation phase using Hebb's rule, and the parameter training phase using STDP and reinforcement learning, so as to form an SNN-based associative memory system. In contrary to just training synaptic weights, Wang et al. propose training both the synaptic weights and delays using gradient descent so as to achieve better performance. Based on a structurally fixed small SNN with sparse recurrent connections, Ponghiran et al. use Q-learning to train only its output layer so as to achieve human-level performance on complex reinforcement learning tasks such as Atari games. Their research demonstrates that a small random recurrent SNN is able to provide a computationally efficient alternative to state-of-art deep reinforcement learning networks with several layers of trainable parameters. The above works have made good progress toward better performing SNN learning algorithms. We believe that further progress will be made in this field in the future. In recent years, hardware oriented DNN compression techniques have been proposed that offer significant memory saving and hardware acceleration (Han et al., 2015a, 2016; Zhang et al., 2016; Huang et al., 2017; Aimar et al., 2018). At present, many compression techniques are proposed that provide a trade-off between processing efficiency and application accuracy (Han et al., 2015b; Novikov et al., 2015; Zhou et al., 2016). Such an approach has also caught on in the design of SNN accelerators (Deng et al., 2019), with the following contribution in this Research Topic. Afshar et al. investigate how a hardware-efficient variant of STDP may be used for event-based feature extraction. Using a rigorous testing framework, a range of spatio-temporal kernels with different surface decaying methods, decay functions, receptive field sizes, feature numbers, and backend classifiers are evaluated. This detailed investigation provides useful insight and heuristics with regards to the trade-off between performance and complexity while using the STDP rule. Pedroni et al. study the impact of different arrangements of synaptic connectivity tables on weight storage and STDP updates for large-scale neuromorphic systems. Based on their analysis, they present an alternative formulation of STDP via a delayed causal update mechanism that permits efficient weight storage and access for both full and sparse connectivity. Other than model complexity, several other papers focus on direct compression of SNNs. Soures and Kudithipudi propose Deep-LSM, a combination of randomly connected hidden layers and unsupervised winner-take-all layers to capture network dynamics followed by an attention modulated readout layer for classification. The connections between hidden layers and winner-take-all layers are partially trained using STDP. Their SNN model is applied in a first-person video activity recognition task, achieving state-of-the-art performance with >90% memory and operation saving compared to the long-short term memory (LSTM). Based on a single fully-connected layer with the STDP learning rule, Shi et al. propose a soft-pruning method that sets a fraction of the weights to the lower bound during training, effectively achieving >75% pruning. Srinivasan and Roy implement spiking convolutional layers comprising of binary weight kernels which are trained using probabilistic STDP, as well as non-spiking fully-connected layers which are trained using gradient descent. A residual convolutional SNN is proposed, which achieves >20x model compression. Programming Tools have been one of the key components driving development in ANN research, examples of which include Theano (Al-Rfou et al., 2016), TensorFlow (Abadi et al., 2016), Caffe (Jia et al., 2014) and Pytorch (Paszke et al., 2019), MXNet (Chen et al., 2015), Keras (Chollet, 2015). These user-friendly programming tools enable researchers to build and train large-scale ANNs using only basic programming know-how. In comparison, the programming tools for SNNs are rather limited. To the best of our knowledge, only SpiNNaker (Furber et al., 2014), BindsNET (Hazan et al., 2018), and PyNN (Davison et al., 2009) provide a basic programming interface to support simple and small SNN simulations. Generally researchers have to build an SNN from the ground up which can be time-consuming and require significantly more programming know-how. Thus, developing user-friendly programming tools to efficiently deploy a large scale SNN is imperative to the advancement of our field. In this Research Topic, an open-source high-speed SNN simulation framework based on PyTorch has been proposed. SpykeTorch (Mozafari et al.) simulates convolutional SNNs with at most one spike per neuron (rank-order coding scheme), and STDP-based learning rules. Although programming tools for SNNs are still in their infancy, we believe that more research needs to be done so that the training of SNNs may approach the efficiency of training ANNs. Recent advances made in modeling SNNs in-silico, as demonstrated by Neurogrid of Stanford University (Benjamin et al., 2014), BrainScales of Heidelberg University (Schemmel et al., 2012), SpiNNaker of University of Manchester, Tianjic of Tsinghua University (Pei et al., 2019), IBM's TrueNorth (Akopyan et al., 2015), and Intel's Loihi (Davies et al., 2018), attest to the great potential of hardware implementation of SNNs. In this Research Topic, Shukla et al. re-model large-scale CNNs so as to mitigate hardware constraints and implement them on the IBM TrueNorth. A CNN used for car detection and counting was demonstrated, with reasonable accuracy compared to a GPU trained CNN but with much lower energy consumption. Bohnstingl et al. implement a learning-to-learn SNN on a neuromorphic chip which accelerates the learning process by extracting abstract knowledge from prior experiences. Other than conventional CMOS circuits, emerging devices such as memristors have also been studied in this Research Topic. Guo et al. propose a STDP-based greedy training algorithm for SNNs to reduce weight levels and enhance robustness toward device non-idealities. They demonstrate online learning on a resistive random access memory (RRAM) system with non-ideal behaviors. Fang et al. propose a generalized swarm intelligence model on SNN: the SI-SNN. SNNs are implemented as agents in swarm intelligence with interactive modulation and synchronization. They implement such neural dynamics on a ferroelectric field-effect transistor (FeFET) based hardware platform to solve optimization problems with high performance and efficiency. In conclusion, it is believed that SNNs achieve superior performance in processing complex, sparse, and noisy spatiotemporal information with high power efficiency exploiting neural dynamics in the event-driven regime. Event-driven communication is particularly attractive in enabling energy efficient AI systems with in-memory computing that will play an important role in ubiquitous intelligence. SNN research is ongoing, and much more progress is to be expected in its learning algorithms, benchmarking framework, programming tools, and efficient hardware. Through cross-discipline exchange of ideas and collaborative research, we hope to build a truly energy-efficient and intelligent machine. This Research Topic is but a small step in this direction; we look forward to more disruptive ideas that distinguish SNNs and neuromorphic computing from the mainstream machine learning approaches in the near future. All authors listed have made a substantial, direct and intellectual contribution to the work, and approved it for publication. This work was partially supported by National Key R&D Program of China (No. 2018AAA0102600 and 2018YFE0200200), Beijing Academy of Artificial Intelligence (BAAI), Initiative Scientific Research Program, and a grant from the Institute for Guo Qiang, TsinghuaUniversity, and key scientific technological innovation research project by Ministry of Education, and Tsinghua–Foshan Innovation Special Fund. The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Abadi, M., Barham, P., Chen, J., Chen, Z., Davis, A., Dean, J., et al. (2016). “Tensorflow: a system for large-scale machine learning,” in 12th {USENIX} Symposium on Operating Systems Design and Implementation ({OSDI} 16) (Savannah, GA), 265–283. Google Scholar Aimar, A., Mostafa, H., Calabrese, E., Rios-Navarro, A., Tapiador-Morales, R., Lungu, I.-A., et al. (2018). Nullhop: a flexible convolutional neural network accelerator based on sparse representations of feature maps. IEEE Trans. Neural Netw. Learn. Syst. 30, 644–656. doi: 10.1109/TNNLS.2018.2852335 PubMed Abstract | CrossRef Full Text | Google Scholar Akopyan, F., Sawada, J., Cassidy, A., Alvarez-Icaza, R., Arthur, J., Merolla, P., et al. (2015). Truenorth: design and tool flow of a 65 mw 1 million neuron programmable neurosynaptic chip. IEEE Trans. Comput. Aided Des. Integr. Circuits Syst. 34, 1537–1557. doi: 10.1109/TCAD.2015.2474396 CrossRef Full Text | Google Scholar Al-Rfou, R., Alain, G., Almahairi, A., Angermueller, C., Bahdanau, D., Ballas, N., et al. (2016). Theano: a Python framework for fast computation of mathematical expressions. arXiv [preprint] arXiv: 1605.02688. Google Scholar Benjamin, B. V., Gao, P., McQuinn, E., Choudhary, S., Chandrasekaran, A. R., Bussat, J.-M., et al. (2014). Neurogrid: a mixed-analog-digital multichip system for large-scale neural simulations. Proc. IEEE 102, 699–716. doi: 10.1109/JPROC.2014.2313565 CrossRef Full Text | Google Scholar Bottou, L. (2012). “Stochastic gradient descent tricks,” in Neural Networks: Tricks of the Trade, eds G. Montavon, G. B. Orr, and K-R. Müller (Berlin; Heidelberg: Springer, 421–436. Google Scholar Chen, T., Li, M., Li, Y., Lin, M., Wang, N., Wang, M., et al. (2015). Mxnet: a flexible and efficient machine learning library for heterogeneous distributed systems. arXiv [preprint] arXiv: 1512.01274. Google Scholar Chollet, F. (2015). Keras [Online]. Available online at: https://keras.io. Google Scholar Davies, M., Srinivasa, N., Lin, T.-H., Chinya, G., Cao, Y., Choday, S. H., et al. (2018). Loihi: a neuromorphic manycore processor with on-chip learning. IEEE Micro 38, 82–99. doi: 10.1109/MM.2018.112130359 CrossRef Full Text | Google Scholar Davison, A. P., Brüderle, D., Eppler, J. M., Kremkow, J., Muller, E., Pecevski, D., et al. (2009). PyNN: a common interface for neuronal network simulators. Front. Neuroinform. 2:11. doi: 10.3389/neuro.11.011.2008 PubMed Abstract | CrossRef Full Text | Google Scholar Dean, J., Corrado, G., Monga, R., Chen, K., Devin, M., Mao, M., et al. (2012). “Large scale distributed deep networks,” in Advances in Neural Information Processing Systems (Lake Tahoe, NV), 1223–1231. Google Scholar Deng, L., Wu, Y., Hu, Y., Liang, L., Li, G., Hu, X., et al. (2019). Comprehensive SNN compression using ADMM optimization and activity regularization. arXiv [preprint] arXiv: 1911.00822. Google Scholar Diehl, P. U., and Cook, M. (2015). Unsupervised learning of digit recognition using spike-timing-dependent plasticity. Front. Comput. Neurosci. 9:99. doi: 10.3389/fncom.2015.00099 PubMed Abstract | CrossRef Full Text | Google Scholar Diehl, P. U., Neil, D., Binas, J., Cook, M., Liu, S.-C., and Pfeiffer, M. (2015). “Fast-classifying, high-accuracy spiking deep networks through weight and threshold balancing,” in 2015 International Joint Conference on Neural Networks (IJCNN) (Killarney: IEEE), 1–8. Google Scholar Franosch, J. M. P., Urban, S., and van Hemmen, J. L. (2013). Supervised spike-timing-dependent plasticity: a spatiotemporal neuronal learning rule for function approximation and decisions. Neural Comput. 25, 3113–3130. doi: 10.1162/NECO_a_00520 PubMed Abstract | CrossRef Full Text | Google Scholar Furber, S. B., Galluppi, F., Temple, S., and Plana, L. A. (2014). The spinnaker project. Proc. IEEE 102, 652–665. doi: 10.1109/JPROC.2014.2304638 CrossRef Full Text | Google Scholar Han, S., Liu, X., Mao, H., Pu, J., Pedram, A., Horowitz, M. A., et al. (2016). EIE: efficient inference engine on compressed deep neural network. ACM SIGARCH Comput. Architect. News 44, 243–254. doi: 10.1145/3007787.3001163 CrossRef Full Text | Google Scholar Han, S., Mao, H., and Dally, W. J. (2015a). Deep compression: Compressing deep neural networks with pruning, trained quantization and huffman coding. arXiv [preprint] arXiv: 1510.00149. Google Scholar Han, S., Pool, J., Tran, J., and Dally, W. (2015b). “Learning both weights and connections for efficient neural network,” in Advances in Neural Information Processing Systems, 1135–1143. Google Scholar Hazan, H., Saunders, D. J., Khan, H., Patel, D., Sanghavi, D. T., Siegelmann, H. T., et al. (2018). Bindsnet: a machine learning-oriented spiking neural networks library in python. Front. Neuroinform. 12:89. doi: 10.3389/fninf.2018.00089 PubMed Abstract | CrossRef Full Text | Google Scholar Hecht-Nielsen, R. (1992). “Theory of the backpropagation neural network,”” in Neural Networks for Perception, ed H. Wechsler (Elsevier), 65–93. Google Scholar Huang, H., Ni, L., Wang, K., Wang, Y., and Yu, H. (2017). A highly parallel and energy efficient three-dimensional multilayer CMOS-RRAM accelerator for tensorized neural network. IEEE Trans. Nanotechnol. 17, 645–656. doi: 10.1109/TNANO.2017.2732698 CrossRef Full Text | Google Scholar Ioffe, S., and Szegedy, C. (2015). Batch normalization: accelerating deep network training by reducing internal covariate shift. arXiv [preprint] arXiv:1502.03167. Google Scholar Jia, Y., Shelhamer, E., Donahue, J., Karayev, S., Long, J., Girshick, R., et al. (2014). “Caffe: convolutional architecture for fast feature embedding,” in Proceedings of the 22nd ACM International Conference on Multimedia (Orlando, FL), 675–678. Google Scholar Jouppi, N. P., Young, C., Patil, N., Patterson, D., Agrawal, G., Bajwa, R., et al. (2017). “In-datacenter performance analysis of a tensor processing unit,” in Proceedings of the 44th Annual International Symposium on Computer Architecture (Toronto, ON), 1–12. Google Scholar Kheradpisheh, S. R., Ganjtabesh, M., and Masquelier, T. (2016). Bio-inspired unsupervised learning of visual features leads to robust invariant object recognition. Neurocomputing 205, 382–392. doi: 10.1016/j.neucom.2016.04.029 CrossRef Full Text | Google Scholar Kingma, D. P., and Ba, J. (2014). Adam: a method for stochastic optimization. arXiv [preprint] arXiv:1412.6980. Google Scholar Novikov, A., Podoprikhin, D., Osokin, A., and Vetrov, D. P. (2015). “Tensorizing neural networks,” in Advances in Neural Information Processing Systems (Montreal, QC), 442–450. Google Scholar O'Connor, P., Neil, D., Liu, S.-C., Delbruck, T., and Pfeiffer, M. (2013). Real-time classification and sensor fusion with a spiking deep belief network. Front. Neurosci.7:178. doi: 10.3389/fnins.2013.00178 PubMed Abstract | CrossRef Full Text | Google Scholar Paszke, A., Gross, S., Massa, F., Lerer, A., Bradbury, J., Chanan, G., et al. (2019). “PyTorch: an imperative style, high-performance deep learning library,” in Advances in Neural Information Processing Systems (Vancouver, BC), 8024–8035. Google Scholar Paugam-Moisy, H., Martinez, R., and Bengio, S. (2006). A supervised learning approach based on STDP and polychronization in spiking neuron networks, IDIAP, EPFL. Google Scholar Pei, J., Deng, L., Song, S., Zhao, M., Zhang, Y., Wu, S., et al. (2019). Towards artificial general intelligence with hybrid Tianjic chip architecture. Nature 572, 106–111. doi: 10.1038/s41586-019-1424-8 PubMed Abstract | CrossRef Full Text | Google Scholar Pérez-Carrasco, J. A., Zhao, B., Serrano, C., Acha, B., Serrano-Gotarredona, T., Chen, S., et al. (2013). Mapping from frame-driven to frame-free event-driven vision systems by low-rate rate coding and coincidence processing–application to feedforward ConvNets. IEEE Trans. Pattern Anal. Mach. Intell. 35, 2706–2719. doi: 10.1109/TPAMI.2013.71 PubMed Abstract | CrossRef Full Text | Google Scholar Querlioz, D., Bichler, O., Dollfus, P., and Gamrat, C. (2013). Immunity to device variations in a spiking neural network with memristive nanodevices. IEEE Trans. Nanotechnol. 12, 288–295. doi: 10.1109/TNANO.2013.2250995 CrossRef Full Text | Google Scholar Roy, K., Jaiswal, A., and Panda, P. (2019). Towards spike-based machine intelligence with neuromorphic computing. Nature 575, 607–617. doi: 10.1038/s41586-019-1677-2 PubMed Abstract | CrossRef Full Text | Google Scholar Schemmel, J., Grübl, A., Hartmann, S., Kononov, A., Mayr, C., Meier, K., et al. (2012). “Live demonstration: a scaled-down version of the brainscales wafer-scale neuromorphic system,” in 2012 IEEE International Symposium on Circuits and Systems (Seoul: IEEE), 702–702. Google Scholar Sengupta, A., Ye, Y., Wang, R., Liu, C., and Roy, K. (2019). Going deeper in spiking neural networks: vgg and residual architectures. Front. Neurosci. 13:95. doi: 10.3389/fnins.2019.00095 PubMed Abstract | CrossRef Full Text | Google Scholar Wu, J., Chua, Y., Zhang, M., Yang, Q., Li, G., and Li, H. (2019a). “Deep spiking neural network with spike count based learning rule,” in 2019 International Joint Conference on Neural Networks (IJCNN) (Budapest: IEEE), 1–6. Google Scholar Wu, Y., Deng, L., Li, G., Zhu, J., and Shi, L. (2018). Spatio-temporal backpropagation for training high-performance spiking neural networks. Front. Neurosci. 12:331. doi: 10.3389/fnins.2018.00331 PubMed Abstract | CrossRef Full Text | Google Scholar Wu, Y., Deng, L., Li, G., Zhu, J., Xie, Y., and Shi, L. (2019b). “Direct training for spiking neural networks: faster, larger, better,” in Proceedings of the AAAI Conference on Artificial Intelligence (Honolulu, HI), 1311–1318. Google Scholar Zhang, S., Du, Z., Zhang, L., Lan, H., Liu, S., Li, L., et al. (2016). “Cambricon-x: an accelerator for sparse neural networks,” in 2016 49th Annual IEEE/ACM International Symposium on Microarchitecture (MICRO) (Taipei: IEEE), 1–12. Google Scholar Zhou, S., Wu, Y., Ni, Z., Zhou, X., Wen, H., and Zou, Y. (2016). Dorefa-net: training low bitwidth convolutional neural networks with low bitwidth gradients. arXiv [preprint] arXiv: 1606.06160. Google Scholar Keywords: deep spiking neural networks, SNN learning algorithms, programming framework, SNN benchmarks, neuromorphics Citation: Li G, Deng L, Chua Y, Li P, Neftci EO and Li H (2020) Editorial: Spiking Neural Network Learning, Benchmarking, Programming and Executing. Front. Neurosci. 14:276. doi: 10.3389/fnins.2020.00276 Received: 13 November 2019; Accepted: 10 March 2020; Published: 15 April 2020. Edited by: Reviewed by: Copyright © 2020 Li, Deng, Chua, Li, Neftci and Li. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. *Correspondence: Guoqi Li, [email protected] These authors have contributed equally to this work
, Paola Corsano
Published: 18 April 2018
Frontiers in Psychology, Volume 9; https://doi.org/10.3389/fpsyg.2018.00558

Abstract:
The Internet was born in the United States in the second half of the twentieth century; it was initially used for military purposes but has since become a powerful instrument for nonmilitary use, including the exchange of information all over the world, thanks to the introduction of tools such as the web browser. From the start, the World Wide Web assumed several functions (e.g., recreation, education, and business) but preserved a private dimension. To connect, people needed access to an Internet-connected computer, which represented a separation from real life, or a virtual reality. A video-terminal device helped these people to immerse themselves in salient but virtual images and sounds; this immersion could induce symptoms such as dissociation (Schimmenti and Caretti, 2010). In the 1990s, scientists developed a conceptualization of the misuse of the Internet and of Internet-addiction disorder (IAD) that was coherent with their conception of the Internet as virtual reality. The strongest criterion for distinguishing healthy Internet use from misuse was connection time; this criterion was supported by several empirical studies regarding its relationship with psychopathological symptoms (Young, 1998; Quayle and Taylor, 2003; Musetti et al., 2016, 2017). However, over the last two decades, Internet use has given rise to global sociocultural changes and has had important implications for the functioning of people's minds (Clowes, 2015). Today, digital and connectable tools such as smartphones are powerful, very small, portable, and (thanks to WiFi and cloud technology) able to store a great deal of salient information about people's lives. These tools thus assume the function of an e-memory (electronic memory) by expanding cognitive memory (Clowes, 2015). Virtual reality is no longer synonymous with the Internet, so there is a need to reformulate the conceptualization of the Internet by taking into account its evolution. The extent of digital information in every sphere of people's lives has caused the integration of the Internet into the cognitive tasks people perform in their daily routines, leading to the consideration of the Internet as part of an extended concept of cognition (Smart et al., 2017). The concept of the Internet as a tool to connect to a virtual reality that is separate from the real world is no longer current, so a new concept of the Internet that takes its environmental features into account is needed. This concept is in line with Floridi's (2014) idea of an infosphere that shapes people's reality. The conceptualization of the Internet as an environment rather than as a tool leads to the reformulation of IAD theory. If the Internet is not just a tool to be utilized, the theoretical model of IAD cannot be based on behavior connected to its overuse, misuse, or abuse. Based on this opinion, we present arguments in favor of reconsidering the Internet as an environment rather than as a tool. In the following section, we explore the Internet's role in cognitive ecology, as well as the inadequacy of treating the Internet as a tool and thus of the current Internet-addiction model. One conceptualization that could help explain the idea that the Internet is a superstructure within which people operate is that of cognitive ecology (Smart, 2017), which has been defined as “the multidimensional contexts in which we remember, feel, think, sense, communicate, imagine, and act, often collaboratively, on the fly, and in rich ongoing interaction with our environments” (Tribble and Sutton, 2011, p. 94). Today's society is digital (Lupton, 2015), and the Internet represents the main part of its cognitive ecology. In the theory of situated cognition (Robbins and Aydede, 2009), cognition is embodied (Gallagher, 2005), embedded (Rupert, 2004), extended, and distributed or collective (Smart et al., 2017). These theories reconceptualize cognition; instead of the classical, individualistic and intra-brain conception of cognition, these theories take into account the relationships among the brain, the body, and the environment to determine the functional products of the mind (Smart et al., 2017). Thanks to the Internet's development (in terms of devices, apps, and social platforms), it can be seen as the principal structure of embodied, embedded, extended, and distributed cognition. Proponents of the embodied-cognition thesis claim that extra-neural bodily factors shape the course of cognitive processing (Anderson, 2003; Shapiro, 2007, 2011). Mobile or wearable devices such as smartphones are today part of people's daily engagements, and they allow continuous online access, which shapes the course of their daily activities and interactions (Smart et al., 2017). By contrast, proponents of the embedded-cognition thesis claim that the extra-organismic environment plays a role (although not a constitutive one) in cognitive states and processes (Rupert, 2004), thus reallocating cognition to within biological boundaries (Smart et al., 2017). The Internet can be inserted within this vision of cognition. For example, augmented reality devices (Smart et al., 2017) such as Google Glass can enrich the sensory experience and have repercussions on cognitive processes. Advocates for the extended-cognition thesis claim that cognitive processes supervene on the relation between a cognitive agent and the social environment in which that agent is situated (Smart et al., 2017). Internal (biological) structures and external devices work in a pair relationship in which biological structures can perform the same operations as external factors (see Clark and Chalmers, 1998) or in a complementary relationship in which external devices can perform operations that biological structures cannot, and vice-versa (see Sutton, 2010; Heersmink, 2015, 2016). The debate regarding the parity or complementarity of the Internet and the brain has not yet been resolved (Smart et al., 2017), and it is not our aim to discuss that issue here. What is important in this context is that Internet devices are so widespread in the social environment that they are the principal external factor through which people's brains relate to and structure external representations; these devices have thus become integrated in people's cognitive architectures (Halpin et al., 2010). Consider the examples of how the use of GPS has modified people's spatial navigation, including its important impact on the neural mechanisms of spatial cognition (Maguire et al., 2000), or considering how Facebook use shapes the representation of the self, including an important impact on the self-concept. This effect is not merely about the interaction between a cognitive agent and environmental devices or about the scaffolding function that external factors have within the mind. The Internet is more than just a scaffold that guides and integrates the mind as it performs functions that the mind cannot accomplish alone (Sterelny, 2010). Rather, people created the Internet to meet people's needs, and the Internet's functions, such as that of e-memory, have changed the ways in which people remember and behave in the world (i.e., a person can recover remote information without having to store every piece of information from day to day). The Internet has changed people's brain structures, which have in turn evolved in such a way as to change how the Internet meets new needs (Clowes, 2013). This view requires consideration of the Internet as an extended function of the mind, including its actual effects on the development of the brain's circuits. In a similar vein, the advent of cooked food changed not only people's tastes but also their digestive functions and the structures of their jaws and teeth; it thus had repercussions on environmental adaptation and species conservation (Wrangham, 2009; Sterelny, 2010). The last thesis regarding the Internet's crucial role is that of distributed cognition. This thesis relates to the cognitive processes (e.g., focusing, reasoning, remembering, and problem-solving) that a collection of individuals share. Again, the Internet has allowed people to take advantage of a huge network of geographically distributed individuals who process cognitive operations at the same time and on the same issue. This opportunity boosts collaboration, information exchange, and the coordination of collective efforts and collective decision-making (Chi et al., 2008; Chi, 2009; Smart et al., 2017). These theories of cognition are today a matter of debate. Some authors have preferred one vision over others; others have considered the theories to not be mutually exclusive and to instead by various integrated aspects of cognition. In the article, we want to underline that, irrespective of the vision that one embraces, the Internet represents a fundamental part of cognitive processing. It not only boosts cerebral operations but also shapes, modulates, and changes neurobiological structures, functioning, and development; the Internet is also, in turn, shaped and developed in a process that resembles a spiral of mutual influence toward ever-higher steps of development. In this sense, a view of the Internet as a mere tool to be utilized functionally or dysfunctionally, as in the model of Internet addiction, is reductive in this era. Thus, considering the Internet as a digital environment that encloses and characterizes cognitive processes is more useful for understanding the phenomenon that we are studying. Consider the people of the nineteenth century, who began to deal with great technological changes (due to the Second Industrial Revolution). The invention of the train, for example, represented a substantial change in the connection between long distances and/or in the amount of people or material carried. People also had to learn to use trains by acquiring new behaviors such as buying tickets and waiting for the departure time; these behaviors could be functional or dysfunctional (examples of the latter include buying an expensive ticket or getting on the wrong train). Although the train was intended as an instrument for traveling to a destination, its growth into a global network and its various functions (industrial, civil, and military) fostered the sociocultural revolution of the 1800s. The train changed the way people thought about industry; thus, in the nineteenth century, the bourgeoisie affirmed its power, and science and literature became more liberal. In other words, what began as a mere instrument evolved into an environmental change that people had to adapt to. The example of the train concretely describes the difference between a tool and a sociocultural environment. The dynamics of the person–tool interaction have been thoroughly studied and represent the basis for the strong Vygotskian psychological tradition (Luria and Vygotsky, 1992). According to this tradition, children organize their behavior by learning to use tools or through external stimuli (Vygotsky, 1997). For example, a child might pay attention to a tool and then name the tool; the name of the tool thus becomes a word in the child's internal speech, thus inducing a new step in the child's reasoning and language functions (Bodrova et al., 2011). This explains how the development of higher brain functions is mediated by the utilization of tools, a view that fits well with the thesis of embodied cognition, according to which external tools shape the course of cognitive processing. It also fits with the thesis of scaffolding cognition, according to which external tools drive cognitive functioning. Within the latter conceptualization, the Internet can be seen as a tool through which people interact and whose use shapes the course of their cognitive processing. However, this view is reductive because it does not take into account the extra-brain operations that the Internet can provide but that the brain cannot. For instance, in the scaffolding view, people can interact with a social platform that reminds them of a salient episode that occurred in their past, thus shaping their emotional reactions and/or thoughts. However, in this view, social-platform interaction does not allow for the improvement of memory systems to provide a better ability to remember salient episodes from the past. Rather, the social platform is seen as a context inside which a limited memory system can take advantage of externally stored information, thus optimizing its work and allowing cognitive resources to be delivered to other processes. In other words, although the Internet—at least in its embryonic form, when recreation was the main online activity—was once considered a tool that shaped and mediated cognition and behavior, today, it is considered an environment that characterizes the people of today. To return to the example of the train, at the beginning, it was considered to be a tool for enhancing travel, but after a few decades, it began to shape the environment that characterized people in the industrial era. Interestingly, Floridi (2014) explained how tools, in addition to being utilized to boost behaviors, have also changed the sociocultural fabrics of various eras, thereby marking the evolution of humanity. The use of bronze (starting in 3000 BC) changed the prehistoric world into the Bronze Age. Similarly, today, people are part of an information society (also known as the infosphere) and can access whatever information they lack (e.g., facts about laws, politics, or science), meaning that there are no boundaries between their online and offline lives—a state known as “onlife” (Floridi, 2014). As the reader may have noted, the arguments in favor of considering the Internet as an environment have multiplied and advanced. It is important to underline this vision here because the classical model and the resulting research into IAD are based on an obsolete conceptualization of the Internet as a tool. Over the last three decades, the literature on this phenomenon has been abundant, but scholars have not reached an agreement on which criteria must be focused on when determining the dividing line between pathological or nonpathological Internet use (Musetti et al., 2016). The main models of Internet-related pathologies retrace those of other addictions (Young, 1998). If the theorists of IAD do not consider the Internet to constitute the current information society, they risk pathologizing a normal behavior, similarly to what happened for new addictions (as with new terms such as “shopaholic” or “workaholic”; see, e.g., Billieux et al., 2015). Without the environmental framework of the Internet, the theorization of pathological Internet use is limited to a reductive list of potentially problematic behaviors (Schimmenti, 2017), such as using the Internet for pornography or gambling. It is noteworthy that the DSM-5 does not resolve this impasse, as it does not mention IAD; the only related disorder, online gaming disorder, is inserted in a section regarding diagnoses that require further study (American Psychiatric Association, 2013). The seven symptoms of IAD in the classical model are withdrawal; tolerance; concern over Internet use; heavier or more frequent Internet use than intended; centralized activities to obtain more from the Internet; loss of interest in other social, occupational, and recreational activities; and disregard for the physical or psychological consequences of Internet use (Young, 1998). These criteria must be present for at least 1 year. Clearly, these criteria are not applicable to the vision of the Internet as an environment. If the Internet constitutes the social fabric, it becomes impossible to withdraw from it, making it impossible to be concerned over Internet use; it likewise becomes impossible to focus on obtaining the Internet. In particular, the criterion of “heavier or more frequent use of the Internet than intended” lacks a comparative parameter in the environmental view of the Internet. How much Internet use is normal if the Internet is ingrained in every part of people's lives and also extends their cognition? In the environmental view, considering the amount of time spent online to be a pathological criterion would mean seeing the entire information society as pathological. Moreover, and paradoxically, a rehabilitation treatment based on this criterion would be centered on reduced Internet access, thus limiting the use of extended and collective cognition (Smart et al., 2017), which could have important repercussions with regard to social adaptation that, in turn, would favor an increase in other pathological criteria, such as withdrawal from social occupation or recreation. Our position is that the classical IAD model should be reformulated to match the vision of the Internet as a social environment. First, researchers must determine whether it is actually possible to be addicted to the Internet. In other words, can people become addicted to their social fabrics? Perhaps it is possible for a person to manifest difficulties or abnormalities when adapting to a social environment. In a similar vein, new models should ignore utilization-related criteria and instead focus on the symptoms that indicate social maladaptation, which may resemble manifestations of known symptoms such as dissociation, depression, anxiety, and personality disorder (Musetti et al., 2018). If this new focus were applied, a question would need be raised about what preexisting pathological conditions would predispose a person to have difficulty adapting to an environment (Caplan, 2002). Considering the Internet as the current socio-cognitive environment, a person's preexisting intra-brain features could favor the success or failure of the adaptation process. In an interesting model, scholars have suggested that maladaptive cognitions precede the symptomatology of IAD (Davis, 2001; Taymur et al., 2016), thus underlining the comorbidity of IAD with heterogeneous psychopathological diagnoses (Orsal et al., 2013). A child presenting with an attention disorder will have some difficulty adapting to a school environment and to a social network of peers, and this difficulty will often impair the development of the child's intellectual and other cognitive functions. Similarly, a person who is cognitively poorly equipped could fail to take advantage of the Internet's contextual affordances (Ryding and Kaye, 2017). This could result in the unsuccessful extension and/or distribution of cognition processes, with repercussions for the person's cognitive development and risks of pathological adaptation to the digitized environment. A similar view could be used in studies on the appropriate treatments for cognitively predisposing features and to help explain the adaptation processes. We are in favor of treating the Internet as a social environment in which a cognitive agent exists. Our proposal is that Internet use should not be seen as a mere instrumental action to achieve a goal (and which could be functional or dysfunctional); rather, we propose treating Internet use as an action situated in the digital context, as part of a system with a proper structure and rules. Considering the concept of the Internet as a social environment, the classical IAD model should be reformulated, as its implications are obsolete and misleading when applied to studies on the pathological population or on potential treatments. AM: devised and structured the paper; PC: contribute to development and deep revision of the work, with literature analysis and agreement for final approval of the paper. The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. American Psychiatric Association (2013). Diagnostic and Statistical Manual of Mental Disorders, 5th Edn. Washington, DC: American Psychiatric Association. Anderson, M. L. (2003). Embodied cognition: a field guide. Artif. Intell. 149, 91–130. doi: 10.1016/S0004-3702(03)00054-7 CrossRef Full Text | Google Scholar Billieux, J., Thorens, G., Khazaal, Y., Zullino, D., Achab, S., and Van der Linden, M. (2015). Problematic involvement in online games: a cluster analytic approach. Comput. Human Behav. 43, 242–250. doi: 10.1016/j.chb.2014.10.055 CrossRef Full Text | Google Scholar Bodrova, E., Leong, D. J., and Akhutina, T. V. (2011). When everything new is well-forgotten old: Vygotsky/Luria insights in the development of executive functions. New Dir. Child Adolesc. Dev. 2011, 11–28. doi: 10.1002/cd.301 PubMed Abstract | CrossRef Full Text | Google Scholar Caplan, S. E. (2002). Problematic Internet use and psychosocial well-being: development of a theory-based cognitive–behavioral measurement instrument. Comput. Hum. Behav. 18, 553–575. doi: 10.1016/S0747-5632(02)00004-3 CrossRef Full Text | Google Scholar Chi, E. H. (2009). “Augmented social cognition: using social web technology to enhance the ability of groups to remember, think, and reason,” in 35th SIGMOD International Conference on Management of Data (New York, NY: ACM), 973–984. Google Scholar Chi, E. H., Pirolli, P., Suh, B., Kittur, A., Pendleton, B. A., and Mytkowicz, T. (2008). “Augmented social cognition,” in AAAI Spring Symposium on Social Information Processing (Stanford, CA), 11–17. Google Scholar Clark, A., and Chalmers, D. (1998). The extended mind. Analysis 58, 7–19. doi: 10.1093/analys/58.1.7 CrossRef Full Text | Google Scholar Clowes, R. W. (2013). The cognitive integration of E-memory. Rev. Philos. Psychol. 4, 107–133. doi: 10.1007/s13164-013-0130-y CrossRef Full Text | Google Scholar Clowes, R. W. (2015). Thinking in the cloud: the cognitive incorporation of cloud-based technology. Philos. Technol. 28, 261–296. doi: 10.1007/s13347-014-0153-z CrossRef Full Text | Google Scholar Davis, R. A. (2001). A cognitive-behavioral model of pathological Internet use. Comput. Human Behav. 17, 187–195. doi: 10.1016/S0747-5632(00)00041-8 CrossRef Full Text | Google Scholar Floridi, L. (2014). The 4th Revolution: How the Infosphere Is Reshaping Human Reality. Oxford: Oxford University Press. Google Scholar Gallagher, S. (2005). How the Body Shapes the Mind. Oxford: Oxford University Press. Google Scholar Halpin, H., Clark, A., and Wheeler, M. (2010). “Towards a philosophy of the web: representation, enaction, collective intelligence,” in Web Science Conference (Raleigh, NC), 1–5. Google Scholar Heersmink, R. (2015). Dimensions of integration in embedded and extended cognitive systems. Phenomenol. Cogn. Sci. 14, 577–598. doi: 10.1007/s11097-014-9355-1 CrossRef Full Text | Google Scholar Heersmink, R. (2016). The internet, cognitive enhancement, and the values of cognition. Minds Machines 26, 389–407. doi: 10.1007/s11023-016-9404-3 CrossRef Full Text | Google Scholar Lupton, D. (2015). Digital Sociology. Abingdon: Routledge. Google Scholar Luria, A. R., and Vygotsky, L. S. (1992). Ape, Primitive Man, and Child Essays in the History of Behavior. New York, NY: Harvester Wheatsheaf. Google Scholar Maguire, E. A., Gadian, D. G., Johnsrude, I. S., Good, C. D., Ashburner, J., Frackowiak, R. S., et al. (2000). Navigation-related structural change in the hippocampi of taxi drivers. Proc. Natl. Acad. Sci. U.S.A. 97, 4398–4403. doi: 10.1073/pnas.070039597 PubMed Abstract | CrossRef Full Text | Google Scholar Musetti, A., Cattivelli, R., Giacobbi, M., Zuglian, P., Ceccarini, M., Capelli, F., et al. (2016). Challenges in internet addiction disorder: is a diagnosis feasible or not? Front. Psychol. 7:842. doi: 10.3389/fpsyg.2016.00842 CrossRef Full Text | Google Scholar Musetti, A., Cattivelli, R., Zuglian, P., Terrone, G., Pozzoli, S., Capelli, F., et al. (2017). Internet addiction disorder o internet related psychopathology? [Internet Addiction disorder or Internet Related Psychopathology?]. Giornale Ital. Psicol. 44, 359–382. doi: 10.1421/87345 CrossRef Full Text | Google Scholar Musetti, A., Terrone, G., and Schimmenti, A. (2018). An exploratory study on problematic Internet use predictors: which role for attachment and dissociation? Clin. Neuropsychiatry 15, 35–41. Google Scholar Orsal, O., Unsal, A., and Ozalp, S. S. (2013). Evaluation of Internet addiction and depression among university students. Proc. Soc. Behav. Sci. 82, 445–454. doi: 10.1016/j.sbspro.2013.06.291 CrossRef Full Text | Google Scholar Quayle, E., and Taylor, M. (2003). Model of problematic internet use in people with a sexual interest in children. CyberPsychol. Behav. 6, 93–106. doi: 10.1089/109493103321168009 PubMed Abstract | CrossRef Full Text | Google Scholar Robbins, P., and Aydede, M. (2009). “A short primer on situated cognition,” in The Cambridge Handbook of Situated Cognition, eds P. Robbins and M. Aydede (Cambridge: Cambridge University Press), 3–10. Google Scholar Rupert, R. (2004). Challenges to the hypothesis of extended cognition. J. Philos. 101, 389–428. doi: 10.5840/jphil2004101826 CrossRef Full Text | Google Scholar Ryding, F. C., and Kaye, L. K. (2017). “Internet Addiction”: a conceptual minefield. Int. J. Ment. Health Addict. 16:225. doi: 10.1007/s11469-017-9811-6 PubMed Abstract | CrossRef Full Text | Google Scholar Schimmenti, A. (2017). Psicopatologia e uso problematico di internet. commento a ≪internet addiction disorder o internet related psychopathology?≫ [psychopathology and problematic internet use. commentary on ≪internet addiction disorder or internet related psychopathology?≫] Giornale Ital. Psicol. 44, 1007–1013. doi: 10.1421/88779 CrossRef Full Text | Google Scholar Schimmenti, A., and Caretti, V. (2010). Psychic retreats or psychic pits?: Unbearable states of mind and technological addiction. Psychoanal. Psychol. 27:115. doi: 10.1037/a0019414 CrossRef Full Text | Google Scholar Shapiro, L. A. (2007). The embodied cognition research programme. Philos. Compass 2, 338–346. doi: 10.1111/j.1747-9991.2007.00064.x CrossRef Full Text | Google Scholar Shapiro, L. A. (2011). Embodied Cognition. Abingdon: Routledge. Google Scholar Smart, P. (2017). Situating machine intelligence within the cognitive ecology of the internet. Minds Machines. 27, 357–380. doi: 10.1007/s11023-016-9416-z CrossRef Full Text | Google Scholar Smart, P., Heersmink, R., and Clowes, R. W. (2017). “The cognitive ecology of the Internet,” in Cognition Beyond the Brain, eds S. J. Cowley and F. Vallée-Tourangeau (ChamSpringer), 251–282. Google Scholar Sterelny, K. (2010). Minds: extended or scaffolded?. Phenomenol. Cogn. Sci. 9, 465–481. doi: 10.1007/s11097-010-9174-y CrossRef Full Text | Google Scholar Sutton, J. (2010). “Exograms and interdisciplinarity: history, the extended mind, and the civilizing process,” in The Extended Mind, ed R. Menary (Cambridge, MA: MIT Press), 189–225. Taymur, I., Budak, E., Demirci, H., Akdag, H. A., Güngör, B. B., and Özdel, K. (2016). A study of the relationship between internet addiction, psychopathology and dysfunctional beliefs. Comput. Human Behav. 61, 532–536. doi: 10.1016/j.chb.2016.03.043 CrossRef Full Text | Google Scholar Tribble, E., and Sutton, J. (2011). Cognitive ecology as a framework for Shakespearean studies. Shakespeare Stud. 39, 94–103. Google Scholar Vygotsky, L. S. (1997). “The history of the development of higher mental functions,” in The Collected Works of L. S. Vygotsky: Vol. 4, eds R. Rieber and A. Carton (New York, NY: Plenum Press), 1–26. Wrangham, R. (2009). Catching fire: How Cooking Made us Human. London: Profile. Young, K. S. (1998). Internet addiction: the emergence of a new clinical disorder. Cyberpsychol. Behav. 1, 237–244. doi: 10.1089/cpb.1998.1.237 CrossRef Full Text | Google Scholar Keywords: digital environment, internet addiction, internet use, cognitive ecology, virtual reality Citation: Musetti A and Corsano P (2018) The Internet Is Not a Tool: Reappraising the Model for Internet-Addiction Disorder Based on the Constraints and Opportunities of the Digital Environment. Front. Psychol. 9:558. doi: 10.3389/fpsyg.2018.00558 Received: 09 October 2017; Accepted: 03 April 2018; Published: 18 April 2018. Edited by: Reviewed by: Copyright © 2018 Musetti and Corsano. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. *Correspondence: Alessandro Musetti, [email protected]
Published: 1 September 2019
Journal: Magyar Tudomány
Abstract:
Hálózati struktúra és centralitás a diplomáciai hálózatokban Network Structure and Centrality and Diplomatic Networks Kacziba Péter PhD, egyetemi adjunktus, Pécsi Tudományegyetem Bölcsészettudományi Kar Politikatudományi és Nemzetközi Tanulmányok Tanszék [email protected] Összefoglalás Hivatkozás Kérjük, válassza ki az önnek megfelelő formátumot: (2017). Magyar Tudomány [Digitális kiadás.] Budapest: Akadémiai Kiadó Letöltve: https://mersz.hu/?xmlazonosito=matud_f28654_p8#matud_f28654_p8 (2019.09.06.) Magyar Tudomány (Budapest: Akadémiai Kiadó, 2017.), letöltve: 2019.09.06. https://mersz.hu/?xmlazonosito=matud_f28654_p8#matud_f28654_p8 BibTeX EndNote Mendeley Zotero Jegyzet elhelyezéséhez, kérjük, lépjen be. Jelen tanulmány célja, hogy multidiszciplináris megközelítéssel vizsgálja az 1817–2015 közötti diplomáciai hálózatok strukturális jellegzetességeit, egyúttal elemezze, hogy a kapcsolatszerkezetben mely államok töltöttek be központi pozíciót. A vizsgálat során negyven hálózati modell elemzéséből következő részeredmények kerülnek bemutatásra: a tanulmány felvázolja a csúcspontokkal, élekkel és azok történeti változásaival kapcsolatos jellemzőket; áttekinti a sűrűséggel és reciprocitással kapcsolatos adatokat; illetve a fokszámcentralitásból kinyert információk alapján megkísérli detektálni a hálózatokban központi szerepet játszó aktorokat. Abstract Hivatkozás Kérjük, válassza ki az önnek megfelelő formátumot: (2017). Magyar Tudomány [Digitális kiadás.] Budapest: Akadémiai Kiadó Letöltve: https://mersz.hu/?xmlazonosito=matud_f28654_p13#matud_f28654_p13 (2019.09.06.) Magyar Tudomány (Budapest: Akadémiai Kiadó, 2017.), letöltve: 2019.09.06. https://mersz.hu/?xmlazonosito=matud_f28654_p13#matud_f28654_p13 BibTeX EndNote Mendeley Zotero Jegyzet elhelyezéséhez, kérjük, lépjen be. This study seeks to analyze the structural characteristics of diplomatic networks between 1817 and 2015 with a multidisciplinary approach and attempt to examine which states have occupied a central position in the network structures. The paper presents partial results of the analysis of forty diplomatic network models: it outlines the topological characteristics of nodes, edges, and their historical transformations; reviews data on density and reciprocity; and on the basis of information obtained from degree centricity, it attempts to detect central actors of the network structures. Hivatkozás Kérjük, válassza ki az önnek megfelelő formátumot: (2017). Magyar Tudomány [Digitális kiadás.] Budapest: Akadémiai Kiadó Letöltve: https://mersz.hu/?xmlazonosito=matud_f28654_p15#matud_f28654_p15 (2019.09.06.) Magyar Tudomány (Budapest: Akadémiai Kiadó, 2017.), letöltve: 2019.09.06. https://mersz.hu/?xmlazonosito=matud_f28654_p15#matud_f28654_p15 BibTeX EndNote Mendeley Zotero Jegyzet elhelyezéséhez, kérjük, lépjen be. Kulcsszavak: nemzetközi kapcsolatok, diplomácia, hálózatkutatás, digitális bölcsészet Hivatkozás Kérjük, válassza ki az önnek megfelelő formátumot: (2017). Magyar Tudomány [Digitális kiadás.] Budapest: Akadémiai Kiadó Letöltve: https://mersz.hu/?xmlazonosito=matud_f28654_p17#matud_f28654_p17 (2019.09.06.) Magyar Tudomány (Budapest: Akadémiai Kiadó, 2017.), letöltve: 2019.09.06. https://mersz.hu/?xmlazonosito=matud_f28654_p17#matud_f28654_p17 BibTeX EndNote Mendeley Zotero Jegyzet elhelyezéséhez, kérjük, lépjen be. Keywords: international relations, diplomacy, network research, digital humanities DOI: 10.1556/2065.180.2019.9.10 Hivatkozás Kérjük, válassza ki az önnek megfelelő formátumot: (2017). Magyar Tudomány [Digitális kiadás.] Budapest: Akadémiai Kiadó Letöltve: https://mersz.hu/?xmlazonosito=matud_f28654_p21#matud_f28654_p21 (2019.09.06.) Magyar Tudomány (Budapest: Akadémiai Kiadó, 2017.), letöltve: 2019.09.06. https://mersz.hu/?xmlazonosito=matud_f28654_p21#matud_f28654_p21 BibTeX EndNote Mendeley Zotero Jegyzet elhelyezéséhez, kérjük, lépjen be. Cikk letöltése Bevezetés Hivatkozás Kérjük, válassza ki az önnek megfelelő formátumot: (2017). Magyar Tudomány [Digitális kiadás.] Budapest: Akadémiai Kiadó Letöltve: https://mersz.hu/?xmlazonosito=matud_f28654_p26#matud_f28654_p26 (2019.09.06.) Magyar Tudomány (Budapest: Akadémiai Kiadó, 2017.), letöltve: 2019.09.06. https://mersz.hu/?xmlazonosito=matud_f28654_p26#matud_f28654_p26 BibTeX EndNote Mendeley Zotero Jegyzet elhelyezéséhez, kérjük, lépjen be. Az előző évtizedekben a globális tér és az abban lezajló interakciók fokozódó komplexitása kihívás elé állította a nemzetközi kapcsolatokkal foglalkozó kutatásokat. Az állami és nem állami szereplők számának drasztikus növekedése, valamint a technikai innovációk és a globalizáció következményei új módszertani alternatívák kidolgozását és átvételét követelték meg. Ezek közül az egyik legígéretesebb lehetőséget a hálózattudomány alkalmazása biztosította, amely fogékony – ennek ellenére periferikus – módszertani alternatívának bizonyult a nemzetközi kapcsolatok fejlődésének lekövetésében. Ugyan a hálózati megközelítés diszciplináris alkalmazása korántsem tekinthető új jelenségnek, a kapcsolatszerkezeti modellezés és elemzés az informatikai, illetve digitális technikák széles körű elterjedéséig ezen a tudományterületen periferikus módszertani alternatíva maradt. A nemzetközi tanulmányokban alkalmazott hálózati módszertan az 1960-as évek óta kiemelt figyelmet szentel a rendszerszintű, globális összefüggések jellemzőivel foglalkozó kutatásoknak. Ezek a strukturális elemzések részletesen vizsgálták az államközpontú nemzetközi rendszer sajátosságait; a kereskedelmi folyamatok hálózati összefüggéseit; a nemzetközi interakciók, tranzakciók és kohéziók tulajdonságait; a szervezeti tagságok következményeit; a regionális klaszterezettséget vagy a szövetségi rendszerek alapvetéseit (Victor et al., 2017, 11.). Utóbbiak mellett a rendszerszintű megközelítések egyik leghangsúlyosabb vonulata az államok közötti diplomáciai kapcsolatokra fókuszált, a diplomáciai linkek meglétéből, hiányából és jellegéből próbált a nemzetközi rendszerre, a centrális vagy periferikus, illetve a regionális vagy individuális szereplőkre vonatkozó tulajdonságokat kinyerni (például: Russet–Lamb, 1969; Neumayer, 2008, Westerwinter, 2017). Hivatkozás Kérjük, válassza ki az önnek megfelelő formátumot: (2017). Magyar Tudomány [Digitális kiadás.] Budapest: Akadémiai Kiadó Letöltve: https://mersz.hu/?xmlazonosito=matud_f28654_p27#matud_f28654_p27 (2019.09.06.) Magyar Tudomány (Budapest: Akadémiai Kiadó, 2017.), letöltve: 2019.09.06. https://mersz.hu/?xmlazonosito=matud_f28654_p27#matud_f28654_p27 BibTeX EndNote Mendeley Zotero Jegyzet elhelyezéséhez, kérjük, lépjen be. Jelen tanulmány ezt a kutatási trendet követve a globális diplomáciai hálózatok általános szerkezeti tulajdonságainak bemutatására, illetve a kapcsolatszerkezetek centrális szereplőinek detektálására vállalkozik. Bár a kutatási eredmények sokrétűek, ehelyütt terjedelmi korlátok miatt az általános szerkezeti tulajdonságok esetében a csúcspontok és kapcsolati linkek összetételére és változásaira, a sűrűségi és reciprocitási adatokra, illetve a fokszámeloszlására fókuszálok; a centralitásra vonatkozó attribútumokat pedig csak a fokszámközpontiság mérőszámai alapján elemzem. Az írásban röviden összefoglalt kutatási eredmények újdonságát elsősorban a mintavételi eljárás széles idősávja biztosítja: míg a korábbi elemzések rövid periódusokra vonatkozóan vizsgálták a diplomáciai hálózatok jellegzetességeit, addig jelen kutatás és tanulmány relatíve széles időskálára, az 1817–2015 közötti időszakra helyezi a hangsúlyt. Hivatkozás Kérjük, válassza ki az önnek megfelelő formátumot: (2017). Magyar Tudomány [Digitális kiadás.] Budapest: Akadémiai Kiadó Letöltve: https://mersz.hu/?xmlazonosito=matud_f28654_p28#matud_f28654_p28 (2019.09.06.) Magyar Tudomány (Budapest: Akadémiai Kiadó, 2017.), letöltve: 2019.09.06. https://mersz.hu/?xmlazonosito=matud_f28654_p28#matud_f28654_p28 BibTeX EndNote Mendeley Zotero Jegyzet elhelyezéséhez, kérjük, lépjen be. A tanulmányban található ábrák saját szerkesztés eredményei, amelyek a modellezett hálózatokból kinyert adatok alapján kerültek összeállításra. Helytakarékossági okokból ezen adatok és hálózatok forrásait külön-külön nem tüntettem fel, ehelyütt azonban szükséges jelezni, hogy az összeállított hálózatok és az abból kinyert adatok forrásai: Bayer, 2006; Moyer et al., 2016; Europa Publications, 2016. Módszertan Hivatkozás Kérjük, válassza ki az önnek megfelelő formátumot: (2017). Magyar Tudomány [Digitális kiadás.] Budapest: Akadémiai Kiadó Letöltve: https://mersz.hu/?xmlazonosito=matud_f28654_p33#matud_f28654_p33 (2019.09.06.) Magyar Tudomány (Budapest: Akadémiai Kiadó, 2017.), letöltve: 2019.09.06. https://mersz.hu/?xmlazonosito=matud_f28654_p33#matud_f28654_p33 BibTeX EndNote Mendeley Zotero Jegyzet elhelyezéséhez, kérjük, lépjen be. A tanulmány kiindulópontjául szolgáló kutatás kifejezetten a diplomáciai kapcsolódások felépítésének és változásainak hosszú idősoros leírására vállalkozott, emiatt azonban számos módszertani kihívással találta szemben magát. Ezek közül az első nagyobb problémát az adatfelvétel jelentette, egységes adatbázis hiányában ugyanis az elemzett periódus diplomáciai kapcsolatait csak három, eltérő módszertant alkalmazó forrásból lehetett kinyerni. Ennek megfelelően az 1817 és 1960 közötti hálózatok forrása a Correlates of War projekt legfrissebb, Reşat Bayer által jegyzett adatbázisa (v2006.1.); az 1960–2010 közötti adatok Jonathan D. Moyer és munkatársai Diplometrics elnevezésű adatgyűjtéséből származnak; míg az utóbbiak által még nem tárgyalt 2015-ös év a The Europa World Year Book diplomáciai információinak eredménye (Bayer, 2006; Moyer et al., 2016; Europa Publications, 2016). Hivatkozás Kérjük, válassza ki az önnek megfelelő formátumot: (2017). Magyar Tudomány [Digitális kiadás.] Budapest: Akadémiai Kiadó Letöltve: https://mersz.hu/?xmlazonosito=matud_f28654_p34#matud_f28654_p34 (2019.09.06.) Magyar Tudomány (Budapest: Akadémiai Kiadó, 2017.), letöltve: 2019.09.06. https://mersz.hu/?xmlazonosito=matud_f28654_p34#matud_f28654_p34 BibTeX EndNote Mendeley Zotero Jegyzet elhelyezéséhez, kérjük, lépjen be. Az adatbázisok jellegéből adódóan a modellezett és vizsgált hálózatok az államok közötti diplomáciai kapcsolatokat ábrázolták, a nemzetközi szervezetekhez küldött állandó képviseletek tehát nem képezték a kutatás tárgyát. Szintén az adatforrások eltérő módszertani sajátosságaiból következően a hálózatokban 1960 előtt diplomáciai akkreditáció eredményezett hálózati kapcsolatot, azt követően viszont már rezidens képviselet jelenléte. Mivel a használt adatbázisok képviseleti osztályozása is eltér egymástól, ezért a kutatás külképviseleti missziók eltérő formáit és szintjeit nem osztályozta vagy kategorizálta: a hálózatok az adott mintavételi évben jelen lévő kapcsolatokat nagyköveti, főképviselői 1 , ügyvivői, miniszteri, esetleg egyéb képviselői 2 szinten ábrázolták, ezek között azonban nem tettek különbséget. Az 1960-at megelőző, akkreditációs adatgyűjtés folyományaként a rezidens képviseletek esetén is minden esetben csak maximum egy diplomáciai képviselet került ábrázolásra, tehát az adott országba delegált vezető képviseleten kívül létesített alsóbbrendű missziók külön nem kerültek feltüntetésre. Hivatkozás Kérjük, válassza ki az önnek megfelelő formátumot: (2017). Magyar Tudomány [Digitális kiadás.] Budapest: Akadémiai Kiadó Letöltve: https://mersz.hu/?xmlazonosito=matud_f28654_p35#matud_f28654_p35 (2019.09.06.) Magyar Tudomány (Budapest: Akadémiai Kiadó, 2017.), letöltve: 2019.09.06. https://mersz.hu/?xmlazonosito=matud_f28654_p35#matud_f28654_p35 BibTeX EndNote Mendeley Zotero Jegyzet elhelyezéséhez, kérjük, lépjen be. A kutatás jobbára ötéves mintavételi időszakokban vizsgálta a külképviseleti kapcsolatokat, bizonyos években (például a második világháború alatt és után) a felhasznált adatbázisok ugyanakkor nem szolgáltattak adatokat. 3 Az elemzésben az adatbázisok által kiválasztott állami entitások szerepelnek, azonban a Correlates of War adatgyűjtési eljárásának megfelelően a Diplometrics gyűjtéséből is ki lettek szűrve az önálló szuverenitással nem vagy csak korlátozottan rendelkező államok. Ez például a 2015-ös évet alapul véve azt jelenti, hogy a 193 ENSZ tagsággal rendelkező országon kívül csak olyan államok kerültek be a hálózatba, amelyek ténylegesen és széles körben tudták gyakorolni szuverenitásukat. 4 Szintén fontos leszögezni, hogy a hálózatok izolált részeket nem tartalmaznak, azaz csak azok az országok kerültek bele a hálózatokba, amelyek az adott mintavételi évben legalább egy másik országgal a kritériumoknak megfelelő diplomáciai kapcsolatban álltak. Hivatkozás Kérjük, válassza ki az önnek megfelelő formátumot: (2017). Magyar Tudomány [Digitális kiadás.] Budapest: Akadémiai Kiadó Letöltve: https://mersz.hu/?xmlazonosito=matud_f28654_p36#matud_f28654_p36 (2019.09.06.) Magyar Tudomány (Budapest: Akadémiai Kiadó, 2017.), letöltve: 2019.09.06. https://mersz.hu/?xmlazonosito=matud_f28654_p36#matud_f28654_p36 BibTeX EndNote Mendeley Zotero Jegyzet elhelyezéséhez, kérjük, lépjen be. A kapcsolatok vizuális modellezését a Gephi 0.9.2 verziószámú szoftver végezte el (Bastian et al., 2009). A modellezett hálózatokban az államok (csúcspontok) közötti diplomáciai interakciókat a bejövő és kimenő képviseletek irányított linkjei jelzik, minden egyes diadikus pár egymáshoz maximum két éllel kapcsolódhat, egy bejövő és egy kimenő misszióval (lásd: melléklet). Utóbbiból következően az összeállított kapcsolati szerkezetek irányított hálózatok. Végezetül fontos kiemelni, hogy az adatbázisok saját bevallásuk szerint is tartalmazhatnak anomáliákat, ekkora adatmennyiség esetében a hibalehetőség tehát nem elhanyagolható tényező (Moyer et al., 2016, 7.). Hálózati struktúra Hivatkozás Kérjük, válassza ki az önnek megfelelő formátumot: (2017). Magyar Tudomány [Digitális kiadás.] Budapest: Akadémiai Kiadó Letöltve: https://mersz.hu/?xmlazonosito=matud_f28654_p41#matud_f28654_p41 (2019.09.06.) Magyar Tudomány (Budapest: Akadémiai Kiadó, 2017.), letöltve: 2019.09.06. https://mersz.hu/?xmlazonosito=matud_f28654_p41#matud_f28654_p41 BibTeX EndNote Mendeley Zotero Jegyzet elhelyezéséhez, kérjük, lépjen be. Az elemzés tárgyát képező negyven hálózat összevetésekor az egyik legszembetűnőbb jellegzetesség a szerkezeti komplexitás folyamatos és fokozatos növekedése, amely bizonyos években kisebb-nagyobb mértékben fluktuált ugyan, a mintavételi időszak egészére nézve azonban jellemző. Ez a komplexitás-növekedés következik egyrészt a hálózatban részt vevő államok számának emelkedéséből, másrészt a közöttük lévő interakciók arányának fokozódásából. A kutatás által modellezett diplomáciai hálózatokban részt vevő államok/csúcspontok (N) száma 1817–1914 között 91,3%-kal emelkedett, amit jóval meghaladott az 1914 és 2015 közötti 345,5%-os növekedés (1. ábra). Ebben a vonatkozásban a német egyesítés alatti és utáni, valamint a második világháború alatti időszak annexiói jelentettek számottevő kivételt, mindkét esetben átmeneti visszaesés mutatkozott a hálózatokban részt vevő államok számában. Hivatkozás Kérjük, válassza ki az önnek megfelelő formátumot: (2017). Magyar Tudomány [Digitális kiadás.] Budapest: Akadémiai Kiadó Letöltve: https://mersz.hu/?xmlazonosito=matud_f28654_p43#matud_f28654_p43 (2019.09.06.) Magyar Tudomány (Budapest: Akadémiai Kiadó, 2017.), letöltve: 2019.09.06. https://mersz.hu/?xmlazonosito=matud_f28654_p43#matud_f28654_p43 BibTeX EndNote Mendeley Zotero Jegyzet elhelyezéséhez, kérjük, lépjen be. 1. ábra. Államok/csúcspontok száma (1817–2015) Hivatkozás Kérjük, válassza ki az önnek megfelelő formátumot: (2017). Magyar Tudomány [Digitális kiadás.] Budapest: Akadémiai Kiadó Letöltve: https://mersz.hu/?xmlazonosito=matud_f28654_p46#matud_f28654_p46 (2019.09.06.) Magyar Tudomány (Budapest: Akadémiai Kiadó, 2017.), letöltve: 2019.09.06. https://mersz.hu/?xmlazonosito=matud_f28654_p46#matud_f28654_p46 BibTeX EndNote Mendeley Zotero Jegyzet elhelyezéséhez, kérjük, lépjen be. A hálózati komplexitás növekedésének másik mozgatórugója az interakciók számának emelkedése volt. Az ezeket reprezentáló hálózati élek (L) száma 1817–1914 között 365,7%-kal, 1914–2015 között pedig 895,2%-kal növekedett (2. ábra). Bár ezek a számok kétségkívül jelentős növekedésről tanúskodnak, fontos azonban megjegyezni, hogy a hálózatokban részt vevő államok maximális kapcsolódási lehetőségeiknek (Lmax) csak átlagosan 33,3%-át használták fel (3. ábra). A hálózati sűrűség (D) értéke az elemzett évek során egyetlenegyszer sem haladta meg a D = 0,5-öt 5 , az arány pedig látványosan csökkent a hálózatok méretének növekedésével. Míg a hálózatokban részt vevő államok 1817–1955 között átlag 37,6%-át teljesítették maximális kapcsolódási lehetőségeiknek, addig 1960 és 2015 között – L max 286,06% növekedése mellett – csak 23,4%-át. A sűrűség látványos csökkenésének egyik oka az államok számának folyamatos emelkedése, amely a maximális kapcsolódási lehetőségek arányát megnövelte. Ez a növekedés az államok közötti interakciók számát fokozta, komplexebb hálózatokat hozott létre, ugyanakkor a diplomáciai költségeket is megemelte. Ezt a költségnövekedést csak a gazdaságilag legfejlettebb államok tudták követni, a többség tehát a rendelkezésre álló források korlátozottsága miatt továbbra is csak a legfontosabb partnerekre koncentrált. A sűrűség csökkenése emellett a felhasznált adatbázisok különböző adatgyűjtési technikáiból is ered. Míg az 1960-as adatok feltételrendszere tartalmazza a többszörös akkreditációkat, addig az 1960 utáni hálózatok csak a rezidens képviseleteket modellezik. Az 1960 utáni csökkenés tehát a multiakkreditációk hiányából is következik. Mintavételi eljárástól függetlenül megállapítható ugyanakkor, hogy az államok számának gyors növekedését a diplomáciai képviseletek számai nem tudták lekövetni, tehát a hálózatok tényleges mérete messze elmarad a lehetséges maximális mérettől. Hivatkozás Kérjük, válassza ki az önnek megfelelő formátumot: (2017). Magyar Tudomány [Digitális kiadás.] Budapest: Akadémiai Kiadó Letöltve: https://mersz.hu/?xmlazonosito=matud_f28654_p48#matud_f28654_p48 (2019.09.06.) Magyar Tudomány (Budapest: Akadémiai Kiadó, 2017.), letöltve: 2019.09.06. https://mersz.hu/?xmlazonosito=matud_f28654_p48#matud_f28654_p48 BibTeX EndNote Mendeley Zotero Jegyzet elhelyezéséhez, kérjük, lépjen be. 2. ábra. Hálózati kapcsolatok/élek száma (1817–2015) Hivatkozás Kérjük, válassza ki az önnek megfelelő formátumot: (2017). Magyar Tudomány [Digitális kiadás.] Budapest: Akadémiai Kiadó Letöltve: https://mersz.hu/?xmlazonosito=matud_f28654_p51#matud_f28654_p51 (2019.09.06.) Magyar Tudomány (Budapest: Akadémiai Kiadó, 2017.), letöltve: 2019.09.06. https://mersz.hu/?xmlazonosito=matud_f28654_p51#matud_f28654_p51 BibTeX EndNote Mendeley Zotero Jegyzet elhelyezéséhez, kérjük, lépjen be. 3. ábra. Hálózati sűrűség (1817–2015) Hivatkozás Kérjük, válassza ki az önnek megfelelő formátumot: (2017). Magyar Tudomány [Digitális kiadás.] Budapest: Akadémiai Kiadó Letöltve: https://mersz.hu/?xmlazonosito=matud_f28654_p54#matud_f28654_p54 (2019.09.06.) Magyar Tudomány (Budapest: Akadémiai Kiadó, 2017.), letöltve: 2019.09.06. https://mersz.hu/?xmlazonosito=matud_f28654_p54#matud_f28654_p54 BibTeX EndNote Mendeley Zotero Jegyzet elhelyezéséhez, kérjük, lépjen be. A sűrűség mért értékei természetesen összefüggésben vannak a hálózat részei­nek összekapcsoltságával, azaz azzal ténnyel, hogy nem minden állam küld minden más államba diplomáciai képviseletet, illetve nem minden ország viszonozza a hozzá érkező missziókat, egyes kapcsolatok tehát aszimmetrikusak maradnak. Az elemzett hálózatok érdekessége azonban az, hogy a kapcsolatok jóval szimmetrikusabbak az átlagos 6 irányított hálózatoknál, a vizsgált időszakban a reciprocitás (r) átlagos értéke r ≈ 0,76 volt. 7 A 4. ábrát elemezve kitűnik, hogy a diplomáciai kölcsönösség a 19. század második felében átlépte ezt az r = 0,7 pontos küszöböt, ettől kezdődően mintavételi eljárástól függetlenül végig átlag felett maradt. A kisebb-nagyobb visszaeséseket áttekintve a szimmetrikusság kiterjedt háborúk, gazdasági válságok következtében (1920, 1935, 1950), illetve tömbök közötti ellenségeskedés (1955), továbbá a dekolonizációs folyamatok (1965) eredményeként átmenetileg csökkent. A diplomáciai képviseletek relatíve magas szimmetrikussága ugyanakkor az alacsony sűrűségi adatokkal együtt értelmezendő. Ahogyan korábban láttuk, a hálózatokban részt vevő államok lehetséges kapcsolataik számának csak relatíve kevés hányadát használták ki, ezekre a megvalósuló kapcsolatokra vonatkoznak az imént ismertetett magas reciprocitási adatok. Az alacsony sűrűségi adatokból azonban az is kiütközik, hogy a hálózatokban létrehozható kapcsolatok többsége nem alakul ki, a hálózati szimmetrikusság magas aránya tehát végső soron nemcsak a létrejövő kapcsolatokra, de a nem megvalósuló kapcsolatokra is igaz. Másként megfogalmazva, az elemzett hálózatokban a létező és a nem létező kapcsolatok többségében egyaránt szimmetrikusságot mutatnak: ha létrejött egy kapcsolat, akkor azt többségében viszonozták 8 , ha viszont nem, akkor az egyoldalú kezdeményezés jobbára elmaradt. Hivatkozás Kérjük, válassza ki az önnek megfelelő formátumot: (2017). Magyar Tudomány [Digitális kiadás.] Budapest: Akadémiai Kiadó Letöltve: https://mersz.hu/?xmlazonosito=matud_f28654_p56#matud_f28654_p56 (2019.09.06.) Magyar Tudomány (Budapest: Akadémiai Kiadó, 2017.), letöltve: 2019.09.06. https://mersz.hu/?xmlazonosito=matud_f28654_p56#matud_f28654_p56 BibTeX EndNote Mendeley Zotero Jegyzet elhelyezéséhez, kérjük, lépjen be. 4. ábra. Reciprocitás (1817–2015) Hivatkozás Kérjük, válassza ki az önnek megfelelő formátumot: (2017). Magyar Tudomány [Digitális kiadás.] Budapest: Akadémiai Kiadó Letöltve: https://mersz.hu/?xmlazonosito=matud_f28654_p59#matud_f28654_p59 (2019.09.06.) Magyar Tudomány (Budapest: Akadémiai Kiadó, 2017.), letöltve: 2019.09.06. https://mersz.hu/?xmlazonosito=matud_f28654_p59#matud_f28654_p59 BibTeX EndNote Mendeley Zotero Jegyzet elhelyezéséhez, kérjük, lépjen be. A hálózat alkotóelemeinek, sűrűségének és szimmetrikusságának arányváltozásai természetesen hatást gyakoroltak a hálózatokban létrejövő kapcsolatok mennyiségére is. Utóbbiak közül az új alkotóelemek megjelenése rövid távon csökkentette a hálózatok sűrűségét, ugyanakkor hosszú távon növelte a létrehozható kapcsolatok számát. Az újonnan belépők kapcsolatrendszerének folyamatos növekedését a kapcsolatok szimmetrikusságra való törekvése is felerősítette, amely – mintegy hálózati befolyásként – a létrejövő egyirányú kapcsolatok többségét idővel szimmetrikussá tette. Bár a hálózatba belépők nagy száma miatt a hálózatok sűrűsége soha nem lépte át a D = 0,5-ös értéket, azonban a fentebb részletezett folyamatok révén az alkotóelemek és a hálózatok fokszáma (k) folyamatosan nőtt. Ehelyütt egyelőre csak az átlagos fokszámot vizsgálva (‹k›) kijelenthető, hogy ez az érték 1817–1914 között 143,4%-kal, 1914 és 2015 között pedig 123,4%-kal növekedett (5. ábra). A fokszám mennyisége természetesen a hálózat különböző alkotóelemeit vizsgálva eltérő módon változott, kijelenthető ugyanakkor, hogy a középértéket a centrális szereplők rendkívül magas értéke jelentősen befolyásolta. Ennek mértéke az idő előrehaladtával folyamatosan nőtt: a legmagasabb fokszámmal rendelkező államok 1817-ben 22,4; 1914-ben 61,1; 1955-ben 121,6; 2015-ben pedig már 307,2 kapcsolattal haladták meg az átlagos fokszám értékét. A centrális és periferikus szereplők kapcsolatainak sűrűsége között tehát jelentős eltérések mutatkoznak (Duque, 2017, 13.). Hivatkozás Kérjük, válassza ki az önnek megfelelő formátumot: (2017). Magyar Tudomány [Digitális kiadás.] Budapest: Akadémiai Kiadó Letöltve: https://mersz.hu/?xmlazonosito=matud_f28654_p61#matud_f28654_p61 (2019.09.06.) Magyar Tudomány (Budapest: Akadémiai Kiadó, 2017.), letöltve: 2019.09.06. https://mersz.hu/?xmlazonosito=matud_f28654_p61#matud_f28654_p61 BibTeX EndNote Mendeley Zotero Jegyzet elhelyezéséhez, kérjük, lépjen be. 5. ábra. Átlagos fokszámeloszlás (1817–2015) Centrális szereplők Hivatkozás Kérjük, válassza ki az önnek megfelelő formátumot: (2017). Magyar Tudomány [Digitális kiadás.] Budapest: Akadémiai Kiadó Letöltve: https://mersz.hu/?xmlazonosito=matud_f28654_p67#matud_f28654_p67 (2019.09.06.) Magyar Tudomány (Budapest: Akadémiai Kiadó, 2017.), letöltve: 2019.09.06. https://mersz.hu/?xmlazonosito=matud_f28654_p67#matud_f28654_p67 BibTeX EndNote Mendeley Zotero Jegyzet elhelyezéséhez, kérjük, lépjen be. Utóbbiból is jól kivehető, hogy a centrális szereplők hálózati befolyása a vizsgált periódus alatt fokozatosan növekedett. Ennek a folyamatnak a követésében a közeliség, a köztiség és a fokszámcentralitás értékeinek elemzése nyújthat segítséget, jelen tanulmány terjedelmi korlátok miatt ezek közül kizárólag utóbbira koncentrál. A fokszámcentralitás a központiság mérésének legkézenfekvőbb mérőszáma, amely az egyes pontok kapcsolati számát viszonyítja egyrészt más csúcspontok fokszámához, másrészt az összes kapcsolat mennyiségéhez (Barabási, 2016, 63.; Newman, 2010, 168–169.). Mivel az elemzett példák irányított hálózatok, ezért a bejövő (kbe ) és kimenő (kki ) fokszámokat különválasztottuk, előbbi a fogadott külképviseleteket, utóbbi a küldött missziók számát jelöli. A bejövő fokszám tekintetében jól kirajzolódik, hogy az elemzett mintegy kétszáz év alatt pusztán négy ország került olyan előnyös helyzetbe, hogy a legmagasabb bejövő fokszámú pozíciót begyűjtse. A negyven mintavételi évben 47,5%-ban Franciaország, 35%-ban az USA, 2,5%-ban az Egyesült Királyság, 15%-ban pedig több ország (Franciaország, Egyesült Királyság, Németország, USA) együttesen osztozott az adott évre vonatkozó legmagasabb fogadott külképviseleti mennyiségen (6. ábra). Az adatokból szintén kitűnik, hogy míg a francia vezető szerep a második és különösen az első világháború előtti időszakra volt jellemző, addig 1950 után az USA vált a diplomáciai külképviseletek legfontosabb célországává. Hivatkozás Kérjük, válassza ki az önnek megfelelő formátumot: (2017). Magyar Tudomány [Digitális kiadás.] Budapest: Akadémiai Kiadó Letöltve: https://mersz.hu/?xmlazonosito=matud_f28654_p69#matud_f28654_p69 (2019.09.06.) Magyar Tudomány (Budapest: Akadémiai Kiadó, 2017.), letöltve: 2019.09.06. https://mersz.hu/?xmlazonosito=matud_f28654_p69#matud_f28654_p69 BibTeX EndNote Mendeley Zotero Jegyzet elhelyezéséhez, kérjük, lépjen be. 6. ábra. Legmagasabb bejövő fokszámmal rendelkező csúcspontok/államok (1817–2015) Hivatkozás Kérjük, válassza ki az önnek megfelelő formátumot: (2017). Magyar Tudomány [Digitális kiadás.] Budapest: Akadémiai Kiadó Letöltve: https://mersz.hu/?xmlazonosito=matud_f28654_p72#matud_f28654_p72 (2019.09.06.) Magyar Tudomány (Budapest: Akadémiai Kiadó, 2017.), letöltve: 2019.09.06. https://mersz.hu/?xmlazonosito=matud_f28654_p72#matud_f28654_p72 BibTeX EndNote Mendeley Zotero Jegyzet elhelyezéséhez, kérjük, lépjen be. A legmagasabb bejövő fokszámmal rendelkező országok listája kiválóan tükrözi a hálózatelmélet presztízsről kialakított elképzeléseit, és igazolja, hogy a magas bejövő kapcsolatszámmal rendelkező csúcspontokhoz érkező kapcsolatok interakciós befektetésnek tekinthetők. A diplomácia világában ez a tulajdonság az úgynevezett szelekciós kényszerből következik, azaz abból a jellegzetességből, hogy az államok korlátozott anyagi erőforrásaikat kizárólag a számukra legelőnyösebb külképviseletek fenntartásába próbálják fektetni. Brandon J. Kinne álláspontja szerint ez a szelekciós preferencia nem pusztán a diadikus/bilaterális párok viszonyrendszeréből következik: az újonnan létesített képviseletek a hálózati szerkezet révén is determináltak, mivel az államok diplomáciai kötelékeiket más országok kapcsolathálója alapján alakítják ki (Kinne, 2014, 247.). Ez a hálózati hatás egy diplomáciai képviselet létesítésekor érvényesül a költségminimalizálás és feladatmaximalizálás kapcsán, valamint a külügyi döntések közvetett külpolitikai jelzései révén. Másként megfogalmazva, a missziót létrehozó állam a bilaterális viszonyok mellett figyelembe veszi a kirendeltséget fogadó ország diplomáciai hálózatának összetételét is, s a képviselet hatásfokának maximalizálására, illetve a költségek minimalizálására törekedve ezután dönt saját külképviseleti szerkezetéről. Ebből következően a legmagasabb bejövő fokszámmal rendelkező országok a képviselet-létesítés szelekciós kényszere miatt a nemzetközi diplomáciai hierarchia csúcsán helyezkednek el, ők rendelkeznek a legtöbb diplomáciai hozzáféréssel, emiatt befolyásuk számottevő (Maliniak–Plouffe, 2011, 4–6.). Széles körű mozgásterüket jól mutatja, hogy a legmagasabb bejövő fokszámok az 1817–1864 és az 1965–1995 közötti időszakoktól eltekintve végig megközelítették az adott évre vonatkozó lehetséges maximumot, ugyanakkor azt ténylegesen soha nem érték el. Hivatkozás Kérjük, válassza ki az önnek megfelelő formátumot: (2017). Magyar Tudomány [Digitális kiadás.] Budapest: Akadémiai Kiadó Letöltve: https://mersz.hu/?xmlazonosito=matud_f28654_p73#matud_f28654_p73 (2019.09.06.) Magyar Tudomány (Budapest: Akadémiai Kiadó, 2017.), letöltve: 2019.09.06. https://mersz.hu/?xmlazonosito=matud_f28654_p73#matud_f28654_p73 BibTeX EndNote Mendeley Zotero Jegyzet elhelyezéséhez, kérjük, lépjen be. Bár a legmagasabb bejövő fokszámmal rendelkező országok presztízse és hálózati befolyása központi szerepet játszik a centralitás vizsgálatakor, azonban nem elhanyagolhatók azok az államok sem, amelyek a mért legmagasabb értékeket megközelítették. A kutatás ezen centrális szereplők kiválasztása érdekében összegyűjtötte azon államokat is, amelyek a mintavételi években bekerültek a tíz legmagasabb bejövő fokszámmal rendelkező ország közé (7. ábra). Annak ellenére, hogy a csoport szereplőinek bejövő fokszámai között jelentős eltérések mutatkoztak, a top 10-ek tagjai mind a diplomáciai tér olyan aktív és nagy befolyással bíró szereplőinek tekinthetők, akikre az államok többsége hasznot eredményező diplomáciai befektetésként tekintett. A csoport elemzésekor az egyik legszembetűnőbb jellegzetesség az abban részt vevők alacsony száma. A mintavételi időszakban összesen harminckét állam 9 tudott bekerülni a mért évben legtöbb bejövő fokszámmal rendelkező tíz ország közé. Ezek közül a központi szereplők közül húsz állam tízszer vagy annál kevesebbszer került be a csoportba, Belgium, Németország, Olaszország, az USA, valamint Franciaország és az Egyesült Királyság viszont harmincnál is többször, utóbbi kettő pedig mind a negyven mintavételi alkalommal. Az imént említett hat ország központi szerepét jól jellemzi, hogy a teljes periódusban létrejövő 98 347 bejövő kapcsolat 14,12%-a ezekbe az országokba irányult. A mért időszakban legmagasabb diplomáciai presztízzsel rendelkezők tehát egy meglehetősen zárt csoportot alkottak, az abba való bekerülés és bent maradás során pedig nem pusztán a diplomáciai vonzerő számított, de az is, hogy az adott állam mióta volt tagja a nemzetközi rendszernek, abban milyen szerepet töltött be, illetve hogy külső és belső viszonyainak stabilitását meddig tudta megőrizni. Hivatkozás Kérjük, válassza ki az önnek megfelelő formátumot: (2017). Magyar Tudomány [Digitális kiadás.] Budapest: Akadémiai Kiadó Letöltve: https://mersz.hu/?xmlazonosito=matud_f28654_p75#matud_f28654_p75 (2019.09.06.) Magyar Tudomány (Budapest: Akadémiai Kiadó, 2017.), letöltve: 2019.09.06. https://mersz.hu/?xmlazonosito=matud_f28654_p75#matud_f28654_p75 BibTeX EndNote Mendeley Zotero Jegyzet elhelyezéséhez, kérjük, lépjen be. 7. ábra. A 10 legmagasabb bejövő és kimenő fokszámmal rendelkező országok csoportja (1817–2015) Hivatkozás Kérjük, válassza ki az önnek megfelelő formátumot: (2017). Magyar Tudomány [Digitális kiadás.] Budapest: Akadémiai Kiadó Letöltve: https://mersz.hu/?xmlazonosito=matud_f28654_p78#matud_f28654_p78 (2019.09.06.) Magyar Tudomány (Budapest: Akadémiai Kiadó, 2017.), letöltve: 2019.09.06. https://mersz.hu/?xmlazonosito=matud_f28654_p78#matud_f28654_p78 BibTeX EndNote Mendeley Zotero Jegyzet elhelyezéséhez, kérjük, lépjen be. Míg a bejövő fokszám az államok egymással szembeni megítélésére utal, addig a kimenő kapcsolatok az egyes országok diplomáciai befektetéseinek fokmérői. Mivel ez a mérőszám nem a hálózaton belüli alkotóelemek döntéseitől, hanem az individuális szereplők egyéni elhatározásától függ, ezért a legfontosabb államok összetétele a bejövő fokszámhoz képest valamelyest diverzifikáltabb képet mutat. Ez jól látszódik a legmagasabb kimenő fokszámmal rendelkezők összetételében, amelyben a mért időszak 32,5%-ában több ország együttesen osztozott a legmagasabb kimenő fokszámon. Ebből következően az önállóan betöltött pozíciók aránya csökkent: az USA a negyven mintavételi év 25%-ában, Franciaország 22,5%-ban, az Egyesült Királyság 12,5%-ban, Oroszország, Ausztria–Magyarország és Olaszország pedig egyenként 2,5–2,5%-ban tudta magáénak a legtöbb kimenő külképviseleti mennyiséget (8. ábra). A legmagasabb kimenő értékkel rendelkező pozíciók diverzifikáltsága az egyes történeti periódusokat vizsgálva is szembetűnő. Míg a legmagasabb bejövő fokszámú adatok esetében egyértelmű volt Franciaország 19. századi és az USA 20. századi vezető pozíciója, addig a kimenő fokszámcentralitás esetében a hosszabb távú egyoldalú elsőbbség csak 1990-től jellemző. Az USA vezető szerepének fokozatos előtérbe kerülését magyarázza az államok számának második világháború utáni drasztikus növekedése, amelynek hatására a diplomáciai költségek megnőttek, és csak a gazdaságilag legfejlettebb államok tudtak extenzív kapcsolathálózatot fenntartani. A kimenő kapcsolatok centrális szereplőinek maximális fokszámhoz viszonyított kimenő fokszáma a folyamat hatására csökkent, a maximális lehetséges kapcsolatok, valamint a megvalósuló linkek száma között egyre nagyobb eltérések mutatkoztak. A 8. ábrán megfigyelhető folyamatot utóbbiak mellett természetesen az adatgyűjtésből kivett multiakkreditációk hiánya is indukálja. Hivatkozás Kérjük, válassza ki az önnek megfelelő formátumot: (2017). Magyar Tudomány [Digitális kiadás.] Budapest: Akadémiai Kiadó Letöltve: https://mersz.hu/?xmlazonosito=matud_f28654_p80#matud_f28654_p80 (2019.09.06.) Magyar Tudomány (Budapest: Akadémiai Kiadó, 2017.), letöltve: 2019.09.06. https://mersz.hu/?xmlazonosito=matud_f28654_p80#matud_f28654_p80 BibTeX EndNote Mendeley Zotero Jegyzet elhelyezéséhez, kérjük, lépjen be. 8. ábra. Legmagasabb kimenő fokszámmal rendelkező csúcspontok/államok (1817–2015) Hivatkozás Kérjük, válassza ki az önnek megfelelő formátumot: (2017). Magyar Tudomány [Digitális kiadás.] Budapest: Akadémiai Kiadó Letöltve: https://mersz.hu/?xmlazonosito=matud_f28654_p83#matud_f28654_p83 (2019.09.06.) Magyar Tudomány (Budapest: Akadémiai Kiadó, 2017.), letöltve: 2019.09.06. https://mersz.hu/?xmlazonosito=matud_f28654_p83#matud_f28654_p83 BibTeX EndNote Mendeley Zotero Jegyzet elhelyezéséhez, kérjük, lépjen be. Szintén a nagyszámú külképviselet fenntartásával kapcsolatos költségekkel állhat összefüggésben, hogy a magas kimenő fokszámmal rendelkező centrális szereplők száma a bejövő fokszámnál tapasztalható arányoknál is alacsonyabb. A mintavételi időszakban huszonhárom ország került be a tíz legtöbb kimenő fokszámot összesítő diagramba (7. ábra), ezek összetétele pedig egyértelműen jelzi, hogy kiterjedt külképviseleti hálózatot csak gazdaságilag fejlett országok tarthattak fenn. Ezek közül a központi szereplők közül kilenc állam tízszer vagy annál kevesebbszer került be a csoportba, a középmezőnybe nyolc ország sorolható, egyenként tíz és harminc közötti előfordulással. A top 10-ek csoportjában legtovább jelen lévők száma ismételten hat: Németország, Olaszország, Oroszország, az USA, valamint Franciaország és az Egyesült Királyság harmincnál is többször került be a csoportba, utóbbi kettő ismételten mind a negyven mintavételi alkalommal. Belgium középmezőnybe történő visszaesése jelzi az ország bejövő és kimenő kapcsolatai között lévő számottevő különbséget, amit az országba delegált nemzetközi szervezetek generálnak, a bejövő kapcsolatok javára. Oroszország belépése a top 6-ok csoportjába szintén érthető fejlemény, ami jól reprezentálja az orosz külpolitika rezsimváltások során is fennmaradó alapvetéseit. A legmagasabb kimenő fokszámmal rendelkező centrális szereplők tehát a legfontosabb célországokhoz hasonlóan egy rendkívül zárt csoportot alkotnak. Erről a zártságról sokat elmond, hogy utoljára Kína volt képes új csatlakozóként belépni a top 10-ek csoportjába, ettől eltekintve a legfontosabb célországok csoportjának tagjai között 1975 óta, míg a legtöbb képviseletet küldő centrális szereplők között pedig 1985 óta nincs változás. Következtetések Hivatkozás Kérjük, válassza ki az önnek megfelelő formátumot: (2017). Magyar Tudomány [Digitális kiadás.] Budapest: Akadémiai Kiadó Letöltve: https://mersz.hu/?xmlazonosito=matud_f28654_p88#matud_f28654_p88 (2019.09.06.) Magyar Tudomány (Budapest: Akadémiai Kiadó, 2017.), letöltve: 2019.09.06. https://mersz.hu/?xmlazonosito=matud_f28654_p88#matud_f28654_p88 BibTeX EndNote Mendeley Zotero Jegyzet elhelyezéséhez, kérjük, lépjen be. A tanulmány által ismertetett adatok a nemzetközi tanulmányok és a hálózatelméleti kutatások számára is kézzelfogható eredményeket produkáltak. A nemzetközi tanulmányok esetében a kutatás egyebek mellett bizonyította, hogy megfelelő adatgyűjtés esetén rendszerszintű összefüggések is modellezhetők, ezek a kapcsolatszerkezetek elemezhetők, a nemzetközi rendszer rétegzettségei, illetve a hatalom absztrakt jellegzetességei pedig kvantitatív módszerekkel kimutathatók. A kapott eredmények cáfolják a feltételezést mely szerint a digitális kapcsolattartás elterjedése miatt a külképviseleti érdekképviselet elvesztheti jelentőségét: a nagykövetségi posztok aránya 2000 és 2015 között – az államok számának 1,6%os növekedése mellett – kb. 19%-kal növekedett, az utolsó mintavételi évben pedig már több mint kilencezer diplomáciai főképviselet működött a világon. Szintén számottevő trend, hogy az eredmények tanulsága szerint a diplomáciai kapcsolódások kialakulását nem pusztán bilaterális (vagy diadikus) viszonyok határozzák meg, a hálózati szerkezet önmagában is döntően befolyásolja a diplomáciai kapcsolatok formálódását, egyúttal az államok abban kifejtett effektivitását. Az elemzés szintén ráirányította a figyelmet arra, hogy a diplomáciai befektetés legfontosabb forrása a külképviselet-létesítés, amely a magas viszonzási hajlandóság miatt növeli a kezdeményező államok nemzetközi beágyazottságát: mivel a kimenő kapcsolatok nagy arányban eredményeznek bejövő kapcsolatokat, ezért az adott állam centralitása, presztízse és diplomáciai hatékonysága a befektetések révén növekszik. Hivatkozás Kérjük, válassza ki az önnek megfelelő formátumot: (2017). Magyar Tudomány [Digitális kiadás.] Budapest: Akadémiai Kiadó Letöltve: https://mersz.hu/?xmlazonosito=matud_f28654_p89#matud_f28654_p89 (2019.09.06.) Magyar Tudomány (Budapest: Akadémiai Kiadó, 2017.), letöltve: 2019.09.06. https://mersz.hu/?xmlazonosito=matud_f28654_p89#matud_f28654_p89 BibTeX EndNote Mendeley Zotero Jegyzet elhelyezéséhez, kérjük, lépjen be. A nemzetközi tanulmányok tudományspecifikus következtetéseinek további listázása helyett érdemes rámutatni néhány olyan eredményre is, amelyek általános hálózatelméleti jelenségekre reflektálnak. Ezek közül az egyik legszembetűnőbb jelenség, hogy a diplomáciai kapcsolódásokat más valódi hálózatokhoz hasonlóan a növekedés jellemzi, ennek a növekedésnek a dinamikája azonban az utolsó mintavételi években fokozatosan lassult, amely természetesen hatást gyakorolt a topológiára és annak tulajdonságaira is. Ez a hálózatok transzformációját nézve azt jelentette, hogy a növekedés dinamikus időszakában egyes országok hosszabb távú hálózati jelenlétük és népszerűségük okán centrális szereplőkké váltak, a megnövekedett szerep azonban sosem eredményezte óriás komponens vagy központosított (csillag formájú) topológia kialakulását. A hálózatokat inkább egy olyan decentralizált struktúra jellemzi, amelyben egyes csúcspontok központi pozíciója a növekedési időszak dinamikus periódusában növekedett, míg annak lassulása során a linkek számának fokozatos kiegyenlítődése révén csökkent. A topológia változása ugyanakkor sosem alakított ki elosztott hálózati szerkezetet, a népszerűségi és erőforrásbeli különbségek a mintavételi periódusban végig fenntartották a diplomáciai hálózatok hierarchikusságát. Melléklet Hivatkozás Kérjük, válassza ki az önnek megfelelő formátumot: (2017). Magyar Tudomány [Digitális kiadás.] Budapest: Akadémiai Kiadó Letöltve: https://mersz.hu/?xmlazonosito=matud_f28654_p94#matud_f28654_p94 (2019.09.06.) Magyar Tudomány (Budapest: Akadémiai Kiadó, 2017.), letöltve: 2019.09.06. https://mersz.hu/?xmlazonosito=matud_f28654_p94#matud_f28654_p94 BibTeX EndNote Mendeley Zotero Jegyzet elhelyezéséhez, kérjük, lépjen be. Diplomáciai képviselek globális hálózatai Hivatkozás Kérjük, válassza ki az önnek megfelelő formátumot: (2017). Magyar Tudomány [Digitális kiadás.] Budapest: Akadémiai Kiadó Letöltve: https://mersz.hu/?xmlazonosito=matud_f28654_p95#matud_f28654_p95 (2019.09.06.) Magyar Tudomány (Budapest: Akadémiai Kiadó, 2017.), letöltve: 2019.09.06. https://mersz.hu/?xmlazonosito=matud_f28654_p95#matud_f28654_p95 BibTeX EndNote Mendeley Zotero Jegyzet elhelyezéséhez, kérjük, lépjen be. Referenciahálózatok Hivatkozás Kérjük, válassza ki az önnek megfelelő formátumot: (2017). Magyar Tudomány [Digitális kiadás.] Budapest: Akadémiai Kiadó Letöltve: https://mersz.hu/?xmlazonosito=matud_f28654_p97#matud_f28654_p97 (2019.09.06.) Magyar Tudomány (Budapest: Akadémiai Kiadó, 2017.), letöltve: 2019.09.06. https://mersz.hu/?xmlazonosito=matud_f28654_p97#matud_f28654_p97 BibTeX EndNote Mendeley Zotero Jegyzet elhelyezéséhez, kérjük, lépjen be. 1817 Hivatkozás Kérjük, válassza ki az önnek megfelelő formátumot: (2017). Magyar Tudomány [Digitális kiadás.] Budapest: Akadémiai Kiadó Letöltve: https://mersz.hu/?xmlazonosito=matud_f28654_p100#matud_f28654_p100 (2019.09.06.) Magyar Tudomány (Budapest: Akadémiai Kiadó, 2017.), letöltve: 2019.09.06. https://mersz.hu/?xmlazonosito=matud_f28654_p100#matud_f28654_p100 BibTeX EndNote Mendeley Zotero Jegyzet elhelyezéséhez, kérjük, lépjen be. 1914 Hivatkozás Kérjük, válassza ki az önnek megfelelő formátumot: (2017). Magyar Tudomány [Digitális kiadás.] Budapest: Akadémiai Kiadó Letöltve: https://mersz.hu/?xmlazonosito=matud_f28654_p103#matud_f28654_p103 (2019.09.06.) Magyar Tudomány (Budapest: Akadémiai Kiadó, 2017.), letöltve: 2019.09.06. https://mersz.hu/?xmlazonosito=matud_f28654_p103#matud_f28654_p103 BibTeX EndNote Mendeley Zotero Jegyzet elhelyezéséhez, kérjük, lépjen be. 2015 Saját szerkesztés (források: Bayer, 2006; Moyer et al., 2016; Europa Publications, 2016; szoftver: Gephi 0.9.2 201709242018) Hivatkozás Kérjük, válassza ki az önnek megfelelő formátumot: (2017). Magyar Tudomány [Digitális kiadás.] Budapest: Akadémiai Kiadó Letöltve: https://mersz.hu/?xmlazonosito=matud_f28654_p106#matud_f28654_p106 (2019.09.06.) Magyar Tudomány (Budapest: Akadémiai Kiadó, 2017.), letöltve: 2019.09.06. https://mersz.hu/?xmlazonosito=matud_f28654_p106#matud_f28654_p106 BibTeX EndNote Mendeley Zotero Jegyzet elhelyezéséhez, kérjük, lépjen be. A tanulmányhoz kapcsolódó kutatás a Nemzeti Tehetség Program támogatásával valósult meg. Pályázati azonosító: NTP-NFTÖ-18-B-0352 Hivatkozás Kérjük, válassza ki az önnek megfelelő formátumot: (2017). Magyar Tudomány [Digitális kiadás.] Budapest: Akadémiai Kiadó Letöltve: https://mersz.hu/?xmlazonosito=matud_f28654_p108#matud_f28654_p108 (2019.09.06.) Magyar Tudomány (Budapest: Akadémiai Kiadó, 2017.), letöltve: 2019.09.06. https://mersz.hu/?xmlazonosito=matud_f28654_p108#matud_f28654_p108 BibTeX EndNote Mendeley Zotero Jegyzet elhelyezéséhez, kérjük, lépjen be. Projekt címe: A nemzetközi kapcsolatok hálózattudományi megközelítései Irodalom Barabási A.-L. (2016): A hálózatok tudománya. Budapest: Libri Könyvkiadó Bastian, M. – Heymann, S. – Jacomy, M. (2009): Gephi: An Open Source Software for Exploring and Manipulating Networks. International AAAI Conference on Weblogs and Social Media. Bayer, R. (2006): Diplomatic Exchange Dataset, v2006.1. http://correlatesofwar.org (letöltve: 2018. 09. 20.) Duque, M. (2017): Core-Periphery Structure in the International Status Hierarchy. Belfer Center for Science and International Affairs, Harvard Kennedy School, 1–38. https://politicalscience.nd.edu/assets/236649/ (letöltve: 2019. 04. 29.) Europa Publications (2016): The Europa World Year Book. 57th edition. London: Routledge Kinne, B. J. (2014): Dependent Diplomacy: Signaling, Strategy, and Prestige in the Diplomatic Network. International Studies Quarterly, 58, 247–259. DOI: 10.1111/isqu.12047 Maliniak, D. – Plouffe, M. (2011): A Network Approach to the Formation of Diplomatic Ties. WORKING DRAFT, San Diego: University of Chicago, 1–24. Moyer, J. D. – Bohl, D. K. – Turner, S. (2016): Diplometrics: Diplomatic Representation. S. Fre­derick S. Pardee Center for International Futures. link (letöltve: 2018. 09. 20.) Neumayer, E. (2008): Distance, Power and Ideology: Diplomatic Representation in a World of Nation-states. Area, 40, 2, 228–236. DOI: https://doi.org/10.1111/j.1475-4762.2008.00804.x Newman, M. E. J. (2010): Networks: An Introduction. New York: Oxford University Press DOI: 10.1093/acprof:oso/9780199206650.001.0001 Russett, B. M. – Lamb, C. W. (1969): Global Patterns of Diplomatic Exchange 1963–1964. Journal of Peace Research, 6, 1, 37–55. https://doi.org/10.1177/002234336900600104 Victor, J. N. – Montgomery, A. H. – Lubell, M. (eds.) (2017): The Oxford Handbook of Political Networks. Oxford: Oxford University Press. DOI: 10.1093/oxfordhb/9780190228217.001.0001 Westerwinter, O. (2017): Uncertainty, Network Change and Costly Signaling: How the Network of Diplomatic Visits Affects the Initiation of International Conflict. Department of Political Science, University of St. Gallen, 1–46. https://www.alexandria.unisg.ch/252634/ (letöltve: 2019. 04. 29.) 1 A Brit Nemzetközösség egymáshoz delegált, nagyköveti rangú képviselői, angolul: High Commisioner. 2 Például apostoli nuncius. 3 A konkrét mintavételi évek a következők: 1817, 1824, 1827, 1832, 1836, 1840, 1844, 1849, 1854, 1859, 1864, 1869, 1874, 1879, 1884, 1889, 1894, 1899, 1904, 1909, 1914, 1920, 1925, 1930, 1935, 1940, 1950, 1955, 1960, 1965, 1970, 1975, 1980, 1985, 1990, 1995, 2000, 2005, 2010, 2015. 4 Névlegesen: Tajvan, a Szentszék és Koszovó. A kutatási következetesség miatt nem került az elemzett országok közé Palesztina és Észak-Ciprus sem, mivel ezekben az esetekben a szuverenitás tényleges gyakorlása erősen korlátozott. 5 Sűrűségi értékhatár: 0 ≤ D ≤ 1. 6 Példának okáért Mark Newman (2010) a World Wide Web irányított hálózatában mért r ≈ 0,54 értéket már „szokatlanul magasnak” minősítette. 7 Reciprocitási értékhatár: 0 ≤ r ≤ 1. 8 Az előzőekben felvázolt számadatok alapján a mintavételi időszakban a létrejövő kapcsolatok átlagosan 76%-os eséllyel váltak szimmetrikussá. 9 A kutatás az adatok összevetésének effektivitása érdekében egyes államok elnevezéseit a vizsgált időszak alatt nem változtatta, azaz egyes releváns esetekben a megszűnő államot és jogutódját kontinuitásként kezelte. Ilyen eset volt például Németország vagy Oroszország. Előbbi esetében a Német Birodalmat, a Weimari Köztársaságot vagy az NSZK-t egyaránt a Németország elnevezés jelöli, utóbbi esetében pedig az Oroszország megjelölés a cári időszakot és a Szovjetuniót egyaránt takarja. Az adatok megfeleltetése természetesen csak releváns esetekben volt lehetséges, példának okáért Ausztria–Magyarország a felbomlás után érthető okokból már külön államként szerepel.
Page of 1
Articles per Page
by
Show export options
  Select all
Back to Top Top