An extensive study of sentiment analysis techniques: A survey

Abstract

Text mining is one of Artificial Intelligent (AI) technologies that uses Natural Language Processing (NLP) for extracting data from structured text. Sentiment analysis (also known as opinion mining) is one of text mining techniques, which is the computational study of personals' behaviours and opinions. The importance of sentiment analysis increases extensively with the enormous growth of the internet sites such as e-commerce, blogs, social media sites, etc. The aim of the survey is to give full view about the most important sentiment analysis techniques. The paper also represents a major contribution through categorizing the SA techniques in an easy and helpful manner for any researcher through classifying SA based on three essential approaches; machine learning-based, lexicon-based, and hybrid approach. Each approach may contain sub-divisions which in turn represent the techniques of SA classification.