Abstract
This article is not only about the speed of digital life in human history, also about its second aspect, that is, the existence of Internet users who use online content for the purpose of insulting or degrading, as well as collecting comments written on social networks with obscene words. It is said that simple and automated actions can be implemented using machine learning. Today, the prevalence of negative comments on online content further exacerbates the problem. Cyberbullying is known to be one of the threats posed by online content, which, in turn, puts online users in an emotional state and causes certain psychological damage. We are collecting a database of obscene comments posted on social networks and used by the media in Kazakhstan. Analyzing complaints received from many social networks, we noticed that the number of publications with offensive, that is, derogatory comments in online content is increasing every day. The results of our research using machine learning methods will help not only to study the roots of abusive language posted on social networks, but also to distinguish the types of offensive comments and obtain automated data sets.