Twitter Sentiment Analysis

CS365: Artifical Intelligence
Jayant Sharma
Aniruddh Vyas
Mentor: Prof. Amitabha Mukerjee

Abstract

We perform a sentiment analysis on a twitter tweet corpus collected during the period January 2009 to March 2010. Using an extended version of the Profile of Mood States (bipolar) questionnaire, we extract the public mood along six bipolar dimensions(Composed, Agreeable, Elated, Confident, Tired, Confused). The results are then compared with some major social, political and economic events during the same period. Its observed that important events have an immediate bearing on the public mood as gleaned from twitter. We also perform a Granger causality analysis on the mood series thus obtained and the Dow Jones Industrial Average (DJIA) closing values during the same period, to check if the two series are correlated and the mood series might contain predictive information about the DJIA series. We find that one of the mood series is strongly correlated with the DJIA time series.

Links

Project Report
Event Timeline and Correlation
Presentation
Project Proposal
Code
References:
  [1]: Bollen, J., Pepe, A., & Mao, H. (2009). Modeling public mood and emotion: Twitter sentiment and socioeconomic phenomena. arXiv.org, arXiv:0911.1583v0911 [cs.CY] 0919 Nov 2009
  [2]: Bollen, J.; Mao, H.; and Zeng, X.-J. 2010. Twitter mood predicts the stock market. Journal of Computational Science 2(1):18
  [3]: Pepe, A., and Bollen, J. 2008. Between conjecture and memento: shaping a collective emo- tional perception of the future. In Proceedings of the AAAI Spring Symposium on Emotion, Personality, and Social Behaviours
  [4]: (For the Twitter dataset) Z. Cheng, J. Caverlee, and K. Lee. You Are Where You Tweet: A Content-Based Approach to Geo-locating Twitter Users. In Proceeding of the 19th ACM Conference on Information and Knowledge Management (CIKM), Toronto, Oct 2010
Click here for information on a local copy of the Twitter dataset(internal link)
Click here for information on a local copy of the Profile of Mood States(POMS); POMS MHS Product Link