DESIGN AND DEVELOPMENT OF FUZZY LOGIC BASED SENTIMENT ANALYSIS SYSTEM FOR ONLINE REVIEWS & SOCIAL MEDIA POST

Abstract

Sentiment is a person’s frame of mind elicited when he/she confronts a specific topic, person, or entity. Comprehending and understanding humans’ views, beliefs, attitudes, or opinions towards a particular entity is sentiment analysis. This process of sentiment analysis can be automated using computational techniques. Users freely express their opinions on websites and social media platforms, we can use this data for analyzing and extracting the sentiment behind the data by applying the concept of fuzzy logic. Since real-world data is imprecise, vague, and not crisp, therefore fuzzy logic is required to deal with such subjective data. Fuzzy approach is based on the premise that key elements in human thinking are not just numbers but can be approximated to tables of fuzzy sets, or, in other words, classes of objects in which the transition from membership to non-membership is gradual rather than abrupt. Sentiment is basically human emotions, understanding human emotions. We have built brain inspired Sentiment Analysis (SA) framework to help machines emulate human inference of sentiment from natural language. We have developed five methodologies for addressing SA using fuzzy logic techniques. The first objective focuses on natural language words; SA is carried out by using these sentiment bearing words only by applying fuzzy logic. The next objective deals with the creation of key phrases and the computation of fuzzy scores for these phrases to perform SA. In the third objective different neuro-fuzzy networks machine learning models for SA are built. The subsequent objective focuses on social media platforms, its importance, and how social media posts can be analyzed using fuzzy concepts for SA. The last objective is about speech emotion recognition systems, and how emotions and sentiment can be evaluated from speech using various multimodal speech and text cues with fuzzy inferencing.