Research on online supporting community of extreme organization by AI-SNA based method

Recently, the propaganda and implementation of traditional extremist activities are facing various interventions and strikes with the living space squeezed. To survive, extreme organizations are constantly seeking new forms of existence and development. With the coming of Web 2.0, they have been transforming their focus from offline to online - using online social networks such as Twitter to broadcast extremism, plan and implement extremist activities. Among those, disseminating information and increasing influence through online supporting community has become a common and effective method. In order to understand the characteristics and laws of this new kind of extremist activity and provide strategies for an effective intervention and strike, in this paper, we design a novel comprehensive research method, which combines Social Network Analysis (SNA) and Artificial Intelligence (AI) techniques. We take the X terrorist organization on the Twitter as a case study to test the superiority of our method. Specifically, first we established the corpus and made a relevant word embedding picture of the organization. Second, based on the picture, we obtained the propaganda characteristics of X terrorist organization and its cooperation organizations. Besides, we further had the future terrorism activity estimated. Third, we constructed a tweet classifier to identify its supporters and got the follow relationship among them and therefore built an online social network supporting community. Finally, we analyzed the community and got the structure and central members with the finding of scale-free network feature. Results show that the proposed method is valuable in the research and application towards the online extremism and the extreme organization supporting community.