Customer satisfaction surveys in the banking sector based on the analysis of text messages

Abstract

The paper proposes an approach to the study of consumer satisfaction. As a source of satisfaction data is used consumer product reviews on natural language published on the Internet. To measure the satisfaction is suggested using sentiment analysis of consumer reviews. For data processing and analysis used sentiment analysis and retrieval aspects of products using machine learning methods. This approach applied for quantitative and qualitative surveys of customer satisfaction of several Russian banks.