An empirical study on investor sentiment and stock returns based on Text Mining

• 1、School of business administration, Liaoning University of engineering and technology, Huludao City, Liaoning Province, 125105

Abstract： In order to accurately capture the indicators reflecting investor sentiment in the Internet information age, this paper uses the python web crawler technology to crawl the posts in the Oriental Fortune stock bar to get the emotional indicators representing investor sentiment. Combined with the information about the discount rate of closed-end funds, the number of IPOs, the number of IPOs, the number of IPOs, the number of IPOs, the number of IPOs, the number of IPOs, the number of IPOs, the number of IPOs, the number of IPOs, the number The composite investor sentiment index is obtained from the composite investor sentiment index of the number of new accounts and consumer confidence index. The results show that: (1) the composite investor sentiment index including text mining variables has a more significant impact on the stock market returns, and has a stronger explanatory and predictive power on the stock market（ 2) The higher the stock market returns, the more positive the investor sentiment in the future.

Keywords： Finance text mining Emotion analysis Investor sentiment stock market