MODELS AND ALGORITHMS FOR PROCESSING TEXT DATA IN SOCIAL NETWORKS

Abstract

The article explores different algorithms for clustering large volumes of textual data. Existing implementation methods were analyzed and Word2Vec and GloVe algorithms were selected. The initial text data for testing the algorithms was obtained by collecting notes from groups created through Telegram. The obtained results showed that the use of these algorithms allows to estimate the frequency of use and importance of individual words in relation to the context of the studied community. Also, the results of the application of algorithms were compared in the work and a conclusion was drawn about their effectiveness.