Sentiment Analysis in Text as an Effective Approach to Behavior Analysis

Abstract

Some approaches for analyzing behavior in the e-government environment are considered. A method for behavior analysis based on sentiment analysis and a naive Bayes classifier is proposed. It is proposed to analyze behavior using the sentiment analysis in text method to determine which documents about citizens collected in state registers belong to positive or negative sentiment. The proposed approach will provide an opportunity to study the socio-political situation in the country more effectively, and to identify the socio-economic problems of citizens.