Abstract
Abstract The article discusses the methods that ensure the detection of destructive messages in the Internet environment. It is shown that to detect them the approaches based on the analysis of the syntactic patterns; analysis of semantic information included in the text, and correlating it with the text corpus; crowd sourcing methods; identification of the patterns of user behavior in social networks; the consideration of the additional information etc., were applied. Good results can be achieved under certain conditions: an access to traffic social networks and other online news resources, opportunities to organize or to get results from crowd sourcing, etc., when there are the restrictions on these conditions the detection of destructive messages can be performed based on indirect signs. The results of the detection of destructive news in the corpus of news reports in Russian, posted on the websites of the Republic of Kazakhstan are shown.