Abstract
Nowadays natural language processing is evolving due to significant improvements in data access and increased computing power, especially in areas such as healthcare, media, finance and human resources. The purpose of this article is to study the methods of intellectual analysis of texts in determining the style of texts and to create an optimal model. The sources of text data used to determine the style of texts are considered. The body of the text is fully processed and experiments are carried out using the necessary methods. K-neighbors and BERT models were used for text classification