Designing a dictionary of patterns of destructive utterances in the task of identifying destructive information influence

E.P Okhapkina,A.O Iskhakova,V.P Okhapkin,A.Y Iskhakov,V Breskich,A Zheltenkov,Y Dreizis

doi:10.1051/e3sconf/202022403013

Abstract

Due to the high level of tension in modern society, social networks are widely used for destructive management of the information space. This aspect of the use of social networks has become particularly important in the light of events taking place in the world (Hong Kong, Syria, France and Ukraine). According to statistics, about 50% of politicized active groups of social networks are subjects to targeted control actions aimed at spreading negative moods in the political sphere. The escalation of conflicts in society generates the most dangerous type of destructive information influence (DII) that require rapid, large-scale coordination of participants in order to attract new supporters and their organizations. Massive DII on the participants of social networks groups exacerbated the problem of promptly identifying the facts of influence, and created serious prerequisites for the development and improvement of methods and means of identifying DII in social networks. The relevance of this problem is due to the existence of a number of methodological and technological problems in the subject area under consideration, one of them is the lack of patterns of network messages containing elements of DII. In the study, the authors consider an approach to designing a dictionary of patterns of destructive utterances.

Highlights

In the context of a large amount of information presented in social networks, an effective approach to solving the problem of identifying destructive information influence (DII) is to create a system for analyzing network messages
It is worth noting that network messages can include various types of content
Like any text analysis system, a DII detection system should include a set of dictionaries, such as a morphological dictionary, a synonym dictionary, tonal labeled dictionaries, and a

Summary

Introduction

In the context of a large amount of information presented in social networks, an effective approach to solving the problem of identifying DII is to create a system for analyzing network messages. It is worth noting that network messages can include various types of content (text, audio, photos, geotags, etc.). Since the object of message analysis is a verbal expression containing signs of DII, for an adequate conclusion about the utterance, it is necessary to include a dictionary of patterns of destructive utterances in the set of dictionaries. Since such dictionaries are not available in the public domain and are not presented in scientific research, the authors propose an approach to building such a dictionary based on the categorical apparatus for constructing ontologies

An approach to designing a dictionary of destructive utterances

Findings

Conclusion