Abstract
The rapid development of the Internet has led to a significant increase in the number of news sites and social networks that describe various events in the world and society. People actively share their opinions about various events in the world. Manually tracking and analyzing such a volume of information is not possible. So, in this way, the use of algorithms for automatic analysis of texts and user comments is an important feature. Published articles and user comments in most cases are of a certain emotional aspect. This article analyzes texts and user comments of Kazakhstan media space. Sentiment classification is done using machine learning algorithms and convolutional and recurrent neural networks (CNN and RNN). A comparative review of the obtained results was performed after the classification.