SentiReview: Sentiment analysis based on text and emoticons

Abstract

Social networking has gradually become a routine for people to post their opinions, views and comments on any product or person. People share their feelings online in a very informal language. Thus, it is very difficult task to analyze exact sentiments attached with that natural language. Sentiment Analysis is a study of people's attitude, opinions, and emotions to classify whether it is positive, negative or neutral. Use of emoticons on social media has increased rapidly in recent years. Hence, we have focused more on how emoticons play an important role in sentiment analysis. Various factors that affect sentiment analysis are discussed briefly in this paper. Also various issues like sarcasm detection, multilingualism, handling acronyms and slang language, lexical variation and dynamic dictionary handling are discussed.