With the advent of web 2.0 and related technologies, the volume of user-generated social media content (UGSMC) is rapidly growing and likely to increase even more in the near future. Social networking apps such as Twitter, Facebook and Google+ are gaining more popularity as they allow people to share and express their views about happenings, have discussion with different communities, or post messages across the world. Twitter sentiment analysis (TSA) extends any organization‘s ability to capture and study public sentiment towards the social events and commodities related to them in real time. This paper provides a comprehensive survey on techniques and applications of TSA available in the literature. The survey focuses on issues such as pre-processing techniques, feature selection methods, learning models, and performance of each method as a criterion. The survey reveals some of the traditional machine learning (ML) algorithms have been efficiently used to work on Twitter data. In conclusion, the paper cites many promising issues for further research in this domain.
Twitter Sentiment analysis, Twitter data, Machine learning, Natural language processing