Modern technologies have made the internet more accessible, with social networks and online communications witnessing a several-fold increase in usage and popularity. Since it offers such ease and convenience, it has become an indispensable part of modern culture. However, as with all good things, it has also led to a new peril - Cyberbullying, which is essentially bullying an individual via an electronic medium. With a high amount of data being shared on social networks daily, automated cyberbullying detection tools need to be put in place. Several approaches have been made for text-based cyberbullying detection, but other modalities such as images have often been ignored. In this paper, we propose a multimodal Long Short-Term Memory (LSTM) network for cyberbullying detection that captures the interplay between the textual and visual modalities by conditioning the LSTM on nontemporal visual data. We further use this network to create an ensemble model that achieves an F1 score of 0.81, outperforming the current state-of-the-art model by 3.8% on the same dataset.

Command and Control (C2) Systems are complex information systems consisting of humans, integrated hardware, and course of action. Traditionally, a documented user manual is provided to the operators to facilitate them with a strong understanding of the system. However, with increasing number of complex functionalities, it becomes difficult and time-consuming for operators to comprehend the document. This study proposes a conceptual framework of Interactive User Manual (C2IUM), which uses a rule-based chatbot and provides clear instructions in a chat-like interface using Natural Language Processing. The paper outlines three-layered methodology including (i) Dataset creation, (ii) Model building, and (iii) Integration. To verify the applicability of the tool, an experiment has been performed on C2 surveillance system and an accuracy of 82% is obtained. The proposed tool introduces automation and enables better customer support in terms of 24 * 7 accessibility, swift answers, and well-structured responses. The tool is robust and scalable.

In the new age of information revolution, recruiters are getting many job applicants’ profiles from various sources. Recruiters invest considerable time and effort to evaluate and organize this amount of data in semi-structured or unstructured format in the information technologies industry. To understand and summarize job applicants’ profiles, a knowledge graph can help to provide instant screening. This paper proposes the use of a machine learning techniques for the generation of knowledge graph and extractive summarization of job applicants. This will help to have a quick knowledge graph visualization and short summary of relevant information from candidates’ profiles. The results of the study can significantly reduce the effort and time taken to manually screen profiles for matching jobs during a recruitment process.

The article examines the features of renewable energy development at the macro and micro levels in Ukraine. The role of knowledge and innovation management in the field of renewable energy has been defined. The subject of the research is the processes of knowledge and innovation management at renewable energy enterprises. The purpose of the article is to identify key factors in the development of renewable energy and approaches to knowledge and innovation management at renewable energy enterprises. General scientific methods have been used, the main of which are: system analysis – application of a comprehensive two-level approach to the analysis of the role of knowledge in the field of renewable energy and correlation and regression analysis. The following results have been obtained: based on the conducted analysis, the expediency of forming an innovation infrastructure for developing such a high-tech sector as renewable energy has been justified. The application of methods and tools of knowledge and innovation management at renewable energy enterprises has been substantiated. Recommendations concerning the formation and development of a knowledge and innovation management system at renewable energy enterprises in Ukraine have been made, as this system will increase the competitiveness of renewable energy in relation to traditional one. Conclusions: the role and features of knowledge and innovation management in renewable energy enterprises are shown. Prospects for the development of renewable energy in Ukraine on the basis of knowledge and innovation management are identified.

This paper illustrates an automatic seizure detection framework that is based on discrete wavelet transforms (DWT), non-linear and statistical features, and support vector machines (SVMs). Electroencephalogram (EEG) signals possess non-linear and rhythmic properties in different frequency bands. Thus, the non-linear features are widely used to advance epileptic seizure detection models and achieve promising results. This research work aims to consider multiple non-linear features so that if the information is missed by one non-linear measure, it can be captured by another. The non-linear features are further combined with the statistical features as statistical features help get better epileptic seizure classification accuracy. All features are calculated on D(2), D(3), D(4), D(5), and A(5) wavelet sub-bands, then combined into a single vector and classified using SVMs. The intended approach’s accomplishment is assessed with respect to terms sensitivity, specificity and accuracy, tested at the University of Bonn and Neurology and Sleep Centre datasets.

Mini-EUSO is a detector observing the Earth in the ultraviolet band from the International Space Station through a nadir-facing window, transparent to the UV radiation, in the Russian Zvezda module. Mini-EUSO main detector consists in an optical system with two Fresnel lenses and a focal surface composed of an array of 36 Hamamatsu Multi-Anode Photo-Multiplier tubes, for a total of 2304 pixels, with single photon counting sensitivity. The telescope also contains two ancillary cameras, in the near infrared and visible ranges, to complement measurements in these bandwidths. The instrument has a field of view of 44 degrees, a spatial resolution of about 6.3 km on the Earth surface and of about 4.7 km on the ionosphere. The telescope detects UV emissions of cosmic, atmospheric and terrestrial origin on different time scales, from a few 𝜇s upwards. On the fastest timescale of 2.5 𝜇s, Mini-EUSO is able to observe atmospheric phenomena as Transient Luminous Events and in particular the ELVES, which take place when an electromagnetic wave generated by intra-cloud lightning interacts with the ionosphere, ionizing it and producing apparently superluminal expanding rings of several 100 km and lasting about 100 𝜇s. These highly energetic fast events have been observed to be produced in conjunction also with Terrestrial Gamma-Ray Flashes and therefore a detailed study of their characteristics (speed, radius, energy ...) is of crucial importance for the understanding of these phenomena. In this paper we present the observational capabilities of ELVE detection by Mini-EUSO and specifically the reconstruction and study of ELVE characteristics.