This article presents methods for identifying the extremist activities of violent groups and individuals within the Internet. We describe our training and testing datasets in Russian and Tatar, as well as research of Russian extremist text characteristics. This resulted in a formation of a feature set for the extremist texts. The applicability of these features for detection of extremist messages was empirically showed.
PDF at the Institute for Systems Analysis of Russian Academy of Sciences website (in Russian): http://www.isa.ru/proceedings/images/documents/2017-67-3/t-3-17_86-97.pdf
Proceeding of the Institute for Systems Analysis of the Russian Academy of Science journal: http://www.isa.ru/proceedings/index.php?option=com_content&view=article&id=979&Itemid=0&lang=en
Ananeva M. I. Devyatkin D. A, Kobozeva M. V., Smyrnov I. V., Solovyev F. N., Chapovsky A. M. The study of extremist texts features // Works of the Institute of System Analysis of the Russian Academy of Sciences. - 2017. - Vol. 67. - №. 3. – Page 86-97.