A Study of Relationship between Investor Sentiment and Stock Price Based on Text Mining

Abstract

Investor sentiment will have an important influence on the trend of stock prices. There is rich information about stock reviews in the online community and investor sentiment analysis based on text mining has become a hot research topic in the field of Social Science in recent years. This thesis takes stock review in the Sina Finance and Economics Stock Forum as the research object, applying support vector machine (SVM) method to the emotional tendency of stock reviews, and then concluding BSI investor sentiment index based on the results of text mining. Finally, a multiple linear regression model is established, which testify the relationship between investor sentiment and stock price. It is found that investor sentiment index based on text mining can effectively improve the forecast of stock price.