METHODS AND TECHNOLOGIES FOR INTELLECTUALIZATION OF SEARCHING FOR DESTRUCTIVE AND RADICAL CONTENT IN SOCIAL MEDIA: ANALYSIS OF THE CURRENT STATE

Abstract

The purpose of this work is a comparative review of methods and technologies for intellectualizing the search for destructive and radical content in social media. The authors propose a typologization and description of the main applied tasks in this field. The analysis of the sources showed that deep learning technologies and the support vector method are the most popular in the professional environment. The authors describe the existing limitations of the intelligent technologies use in the subject area. As a general result of the conducted analytical research, we note the following. Partially formulated conclusions are characteristic not only for such a subject area as the intellectualization of the search for destructive content. In particular, close attention is paid to the problem of compiling qualitative data samples as hindering the further development of artificial intelligence technologies. As trends gaining relevance in the field of countering destructive content, it is worth highlighting, firstly, the need to form interdisciplinary scientific and technical teams, and, secondly, the shift in emphasis in the classification of content towards video and image processing. Intensive involvement of experts in the field of destructive content research – linguists, sociologists and psychologists – to automate its search and analysis will increase the overall degree of formalization of such definitions as “destructive”, “radical” and “extremist”, as well as their inherent features, which will ensure a transition to a more effective level of algorithmization. The growth trend of two directions is predicted: the development of organizational and managerial methods for the formation of interdisciplinary research teams to solve the identified applied tasks, as well as the development of content classification methods based on video and image processing against the background of a decrease in the importance of text content in social media.