Нейросетевой метод классификации текстов групп социальной сети Вконтакте

Abstract

Classification of text data using neural networks is used in numerous projects. Different structures of artificial neural networks with varying parameters can be used. This article describes the process of creating and training a neural network aimed at the classification of records from the social network Vkontakte. The aim of the work is to check the applicability of neural networks to solve the problem. Tensorflow package was used as a software tool to create a neural network. Data for neural network training are obtained from the records of groups with specific topics (biology, information technology, cinema). The algorithm of automatic of Vkontakte groups texts, tested in the study, will be used in the development of the content component of the functional psychometric model of cognitive – behavioral predictors of personal activity in the framework of its educational activities.