Natural language processing to analyze user feedback to identify trends in large volumes of textual business data

Abstract

The purpose of the study is to develop a system for automated analysis of user feedback based on natural language processing methods to detect tone, key aspects and topics. The object of the study is the processes of processing and analyzing user feedback in a digital environment. The subject of the study is natural language processing methods and models for automating the analysis of text feedback. The research methods include machine learning, deep neural networks, statistical methods of text analysis and methods for assessing the quality of models. Classical algorithms, neural network models and transformer architectures are used.