The temporal evolution of the normalized web distance
- 3 October 2016
- journal article
- Published by Emerald in Internet Research
- Vol. 26 (5), 1269-1290
- https://doi.org/10.1108/intr-07-2014-0185
Abstract
Purpose: The purpose of this paper is to assess whether the temporal evolution of the normalized web distance (NWD) between significant terms concerning, e.g., a case of online activism can be used as a meta-data technique to measure evolution over time of, e.g., progress or decline of social empowerment. Design/methodology/approach: The NWD between two terms has been identified as a quantitative measure for semantic proximity, ascertaining a defining relation between them. A trend analysis is made by performing on the internet a time window restrained series measurement of NWD of all combinations of key-terms and classifier-terms. Case defining key-terms, positive and negative discourse polarizing classifier-terms, and neutral classifier-terms for negative control need to be determined by discourse analysis of information on a targeted case. An example of NWD evolution from 1994 until 2013 is presented to measure the empowerment effects of the Wirikuta online movement on the Huichol people in Mexico. Findings: The application of the NWD temporal evolution method to the Wirikuta case shows a slight but significant semantic change of the key-terms with respect to some of the positive and negative classifier-terms. The neutral classifier correctly shows no significant distance variation, as required for valid application of the method. The method provides indications for a complex image of empowerment of the Huichol identity. Research limitations/implications: The accuracy of the method is limited due to short-term and between-user variability of the search tool’s page counts. More reliable access to a web-index will be required for more accurate NWD-based trend analysis. Practical implications: The monitoring of temporal NWD evolution provides a potential tool for more comprehensive trend description compared to classical frequency based methods. Originality/value: Trend analysis is key to internet research, to which the temporal NWD method provides an innovative contribution.Keywords
This publication has 10 references indexed in Scilit:
- The power of prediction with social mediaInternet Research, 2013
- Forecasting the pulseInternet Research, 2013
- Understanding the predictive power of social mediaInternet Research, 2013
- Predicting information credibility in time-sensitive social mediaInternet Research, 2013
- Critical theory for women empowerment through ICT studiesQualitative Research Journal, 2013
- Hit Count Reliability: How Much Can We Trust Hit Counts?Lecture Notes in Computer Science, 2012
- The New Public Sphere: Global Civil Society, Communication Networks, and Global GovernanceThe Annals of the American Academy of Political and Social Science, 2008
- What's in a Name: Is “Evolutionary Psychology” Eclipsing “Sociobiology” in the Scientific Literature?Evolutionary Psychology, 2007
- Googleology is Bad ScienceComputational Linguistics, 2007
- The Google Similarity DistanceIEEE Transactions on Knowledge and Data Engineering, 2007