An Automatic Rating System Based on Review Sentiments and Intuitionistic Fuzzy Sets

Abstract

Natural language processing (NLP) and text mining methodologies are beneficial tools for utilizing from large scale service review datasets. Sentiment analysis (SA) is among them and it is used to label texts according to their polarity degrees. In this paper, a novel review scoring approach is introduced based on sentiment polarity scores and intuitionistic fuzzy sets. The approach is then implemented on a review dataset which is retrieved from several websites for a certain hotel service. The proposed approach presents promising results for more reliable product/service evaluation relatively to crisp rating scores. © 2023 Elsevier B.V., All rights reserved.