Study of the evaluation function of a sentence with the use of neural networks

The article discusses the problem of obtaining an emotional-evaluative function of feedback about films, books, etc. on the basis of emotionally-colored lexemes it contains. Various techniques have been used to solve this problem, including regression analysis, expert systems, SVM, and many others. We used an approach based on neural networks, a method that has shown its effectiveness in solving problems of classification, forecasting, and pattern identification. The results turned out to be quite encouraging and allow further research in this direction.