Twitter based Sentiment Analysis for Subject Identification

Abstract: Nowadays the social media such as blogs, Twitter, Facebook etc. are widely used for participatory information sharing and collaboration. The analysis of the dynamic opinion of users in these social media forums will helps to answer many questions from various business and research fields across the world. Sentiment mining is used to identify and extract the sentiment and other emotional states in the online text. This paper aims to study the subject identification by analyzing tweets from Twitter, a micro-blogging site.

Keywords: Sentiment mining, Twitter, Naive Bayes, Fisher, API, Bag of Words, Bag of N-Grams.