A FRAMEWORK FOR ASPECT BASED SENTIMENT ANALYSIS USING FUZZY LOGIC

Abstract

Sentiment Analysis (SA) is the study of people’s opinions, emotions, and appraisals toward products and events. In the past years, it fascinated a great deal of attentions from both industry and academia for a variety of applications. Opinions are significant, because people need to make decisions. It is helpful not only for the individuals but also for the business organizations. Fuzzy logic can provide a quick way to solve the haziness present in most of the natural languages. The techniques are less explored in sentiment analysis. In this paper, Aspect based Sentiment Summarization (ASFuL) is proposed with fuzzy logic by classifying opinions polarity as strong positive, positive, negative and strong negative. It also integrates the non-opinionated sentences using Imputation of Missing Sentiment (IMS) mechanism which plays a vital role in generating precise results. The researchers used Fuzzy Logic to find sentiment classes in the review. The results show that the mechanism is viable to extract opinions in an efficient manner.

Keywords

• ASFuL
• Sentiment Analysis
• Aspect
• Sentiment Summarization
• Fuzzy Logic