Автоматизированное обнаружение перекрестных связей пользователей ультраправых сообществ в социальной сети
| Parent link: | Вестник Томского государственного университета. Философия. Социология. Политология/ Национальный исследовательский Томский государственный университет (ТГУ) № 59.— 2021.— [С. 156-166] |
|---|---|
| Autore principale: | |
| Enti autori: | , |
| Altri autori: | , |
| Riassunto: | Заглавие с экрана В статье представлен анализ современных исследований и методов, направленных на автоматизированное обнаружение экстремистских сообществ, в частности -автоматизированное обнаружение перекрестных связей сообществ социальной сети. В качестве перспективного метода автоматизированного обнаружения перекрестных связей пользователей ультрарадикальных сообществ, в частности ультраправых, предлагается использование существующих парсинговых сервисов. In the last decade, the coverage of social networks on the Internet by radical groups has expanded and provided militant extremists with many opportunities to recruit new adherents, build chains of interactions, and distribute illegal content. Extremist organizations engage in targeting, recruiting new members on social sites such as Facebook, VKontakte, and on radicalized web forums, including within individual communities. The main danger of online radicalization lies in its ability to quickly "infect" large online communities with destructive content. On the other hand, the Internet facilitates the study of extremist views. The use of automated or semi-automatic data collection tools based on the analysis of the website text content where extreme opinions are potentially distributed allows identifying incident planning quickly. As part of the study, we formulate an assumption on the effectiveness of using existing parsing services to automate the detection of cross-links between ultra-right community users. In this article, we consider only information that is freely available on the Internet. As an experimental site, we chose the social network VKontakte. To solve the problem, we analyzed the involvement of the community in radical far-right groups, as well as cross-links between them and student communities selected as experimental. Two student communities SC 1 and SC 2 were selected with the number of subscribers 22,000 and 1,600, respectively. The expert selected 30 radical ultra-right communities. According to the proposed algorithm, we found users who are members of a radical far-right community, as well as users who are simultaneously members of several radical far-right communities. The study showed that automated detection of cross-links between users of radical far-right communities in a social network is an achievable goal. However, various parsing services for similar requests provide results different from each other. Thus, the challenge remains to develop software tools for automating sociological research based on data retrieved from social networks. |
| Pubblicazione: |
2021
|
| Soggetti: | |
| Accesso online: | https://doi.org/10.17223/1998863X/59/15 |
| Natura: | Elettronico Capitolo di libro |
| KOHA link: | https://koha.lib.tpu.ru/cgi-bin/koha/opac-detail.pl?biblionumber=664978 |