NEW TECHNOLOGIES TO IDENTIFY ALT-RIGHT EXTREMIST COMMUNITIES IN SOCIAL MEDIA

The article makes an analytical review of approaches to creating ultra-radical communities identifying tools for social media. The formulation of the radicalization process, as a transition from nonviolent forms of expression to violent actions, is indicated. Radicals should be seen as committed to decisive action, which may not be extreme. Extremists should be considered as supporters of using extreme means to achieve their goals, even if it is possible to achieve the goal legally, without using violence. Terrorists are those who use ideologically motivated violence. The measure of the legal assessment of the actions of each category is the assessment of the level of danger of the methods they use for society and the state. The prefix "ultra" means that the radicals accept the idea that violence is necessary to advance an ideological goal. They justify, propagandize this idea, express readiness to proceed to violent actions, and also undertake a moral obligation in defense of groups that promote this idea. Key characteristics and ideological platforms of ultra-right extremist communities, threats and risks of online radicalization are described. The fundamental idea of ultra-right communities is the belief in the superiority of some individuals and groups and the inferiority of others. It emphasizes the importance of software algorithms that allow diagnosing various kinds of threats in the information space with high speed. Creation of new technologies for the identification of extremist communities in social media is possible in interdisciplinary research teams consisting of sociologists, computer science specialists, linguists, psychologists, criminologists. The "background characteristics" and "background analytics" are used as the basis for creating software algorithms. Problems with the creation of research tools and services are indicated, the authors' classification and description of the strengths and weaknesses of approaches to create tools for identifying extremist communities are given. The following categories are indicated: autonomous, additive, hybrid, multi-level. A description of the authors' method of calendar-correlation analysis is proposed for detecting the target ultra-radical group on the characteristic change in the group's activity in the "neighborhood" of key events. Currently, the team of authors is testing the prototype.