Machine learning models for bank reviews classification

Abstract

Using the banking products and services review corpus, analysis is conducted to establish different text classification models. The paper explores different approaches to the processing of unstructured textual information. Based on the selected approaches, the review corpus on banking products and services received during the COVID-19 pandemic is analyzed. An automatic Internet resources parser has been developed to obtain the required training sample. Software has been developed that implemens basic methods for the classification models construction. This model can be used to create system for monitoring people’s attitudes to banking processes.