A Review of Sentiment Analysis Techniques using Soft Computing Approaches

Sentiment analysis and opinion mining have taken a leading position ever since the boom of the internet which led to an enormous amount of opinionated data being posted online. Hence, it has become essential to structure an efficient algorithm for analysing sentiments from text. Even though traditional machine learning algorithms were a breakthrough in the domain, there remained a demand for a better solution. This paper presents a systematic & qualitative literature review for various approaches to sentiment analysis using hybridized soft-computing techniques to cater to noise and improve the flexibility of a fuzzy inference-based sentiment analysis system. This qualitative literature review adds to the body of evidence that soft-computing and fuzzy logic techniques can be effectively applied to a range of applications where noise is inherent, especially in the case of sentiment analysis and opinion mining.