chat users

A method for detecting sources of disinformation and inauthentic behavior of chat users in social networks has been developed. The developed model is based on the analysis of text information using modern machine learning algorithms, in particular classifiers (SVM, Naive Bayes, decision trees, etc.) and clustering methods to identify structural relationships between news and users. Considerable attention is paid to the collection and balancing of datasets, as well as the visualization of networks to assess the spread of fake news.