Topic Modeling based Sentiment Analysis on Social Media for Stock Market Prediction

Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing, pages 1354–1364, Beijing, China, July 26-31, 2015. c⃝2015 Association for Computational Linguistics Topic Modeling based Sentiment Analysis on Social Media for Stock Market Prediction Thien Hai Nguyen Kiyoaki Shirai School of Information Science Japan Advanced Institute of Science and Technology 1-1 Asahidai, Nomi, Ishikawa 923-1292, Japan {nhthien, kshirai}@jaist.ac.jp Abstract The goal of this research is to build a model to predict stock price movement us- ing sentiments on social media. A new feature which captures topics and their sentiments simultaneously is introduced in the prediction model. In addition, a new topic model TSLDA is proposed to obtain this feature. Our method outperformed a model using only historical prices by about 6.07% in accuracy. Furthermore, when comparing to other sentiment anal- ysis methods, the accuracy of our method was also better than LDA and JST based methods by 6.43% and 6.07%. The results show that incorporation of the sentiment information from social media can help to improve the stock prediction. 1 Introduction Stock price forecasting is very important in the planning of business activity. However, building an accurate stock prediction model is still a chal- lenging problem. In addition to historical prices, the current stock market is affected by the mood of society. The overall social mood with respect to a given company might be one of the important variables which affect the stock price of that com- pany. Nowadays, the emergence of online social networks makes large amounts of mood data avail- able. Therefore, incorporating information from social media with the historical prices can improve the predictive ability of the models. The goal of our research is to develop a model to predict a stock price movement using information from social media (Message Board). In our pro- posed method, the model predicts the movement of the stock value at t using features derived from information at t − 1 and t − 2, where t stands for a transaction date. It will be trained by supervised machine learning. Apart from the mood informa- tion, the stock prices are affected by many factors such as microeconomic and macroeconomic fac- tors. However, this research only focuses on how the mood information from social media can be used to predict the stock price movement. That is, the mood of topics in social media is extracted by sentiment analysis. Then, the topics and their sentiments are integrated into the model to pre- dict the stocks. To achieve this goal, discover- ing the topics and sentiments in a large amount of social media is important to get opinions of investors as well as events of companies. How- ever, sentiment analysis on social media is diffi- cult. The text is usually short, contains many mis- spellings, uncommon grammar constructions and so on. In addition, the literature shows conflict- ing results in sentiment analysis for stock market prediction. Some researchers report that the senti- ments …

Content truncated to 500 words