Abstract
This article is devoted to analysis of the articles on cyberbullying of children and adolescents and creating the methodology of working on the own work, which related especially to Kazakh social media. The article provides examples of similar works by foreign researchers, the United States, England and other European countries. Using the methods of theoretical, namely analysis, synthesis, and empirical: comparison and experiment, the analysis of works on this topic was carried out, and this work was aimed at analyzing the problem in the Kazakh space. In brief, we consider the issue of creating a parser, as well as collecting data for training machine learning and deep learning algorithms that could find and block texts containing a humiliating slope in real time. The article will be of interest to both novice specialists who are engaged in data analysis, and experienced ones to expand their horizons.The question of implementing deep learning algorithms requires further study, which is one of the parts of a large work being done, as well as a future topic for a subsequent article.