Research On The Correlation Between Investor's Sentiment And Sse 50 Index Based On Text Mining

China's stock market has experienced only 30 years of vicissitudes since it was born.However,the U.S.stock market has already reach it mature period.The new year is approaching,the number of investors in China's market has exceeded 180 million,and the number of participants is large.Scholars and investors at home and abroad are flocking to it.There are many factors that affect the fluctuation of stock market index,including direct factors and potential factors that are difficult to detect.In recent years,with the continuous enhancement of computer computing ability and the emergence of artificial intelligence algorithms,more and more scholars and investors have carried out a series of research on stock price fluctuations by combining artificial intelligence with stock market rules.Based on the comments of retail investors on the Internet,this paper explores the emotional tendencies in language information and constructs an emotional index model based on the comments of retail investorsin in order to studies the correlation between it and the SSE 50 Index.In order to apply the irrational decision-making factors of investors to the study of the relationship between the Shanghai Stock Exchange 50 Index,this paper combines machine learning algorithm to build a representative investor sentiment index model,and proposes a sentiment tendency extraction and classification method based on the combination of natural language processing and naive Bayes model.A more stable emotional score is calculated,after noise reduction,a series of correlation analysis is carried out with the stock index changes.The results are as follows(1)In the construction of investor sentiment index,we should not only consider positive indicators such as comments and likes,just as the number of users' fans selected in this paper is used as the authoritative indicator to be included in the measurement standard.That build a composite sentiment measurement expression,which can make the sentiment measurement expression more representative.(2)From the overall trend,the trend of the sentiment index is roughly similar as that of SSE 50 Index,but the investment sentiment of individual investors fluctuates relatively large,which makes it easier to catch up and kill down.When the stock index is at a high level,it shows "excitement",and when it is at a low level,it shows "panic",which is consistent with the reality.It shows that this paper constructs a relatively effective emotional index model.(3)Under the premise of big data,MIC performs better in the correlation analysis of sentiment and stock index.MIC is more suitable for the correlation analysis of big data.At the same time,as results of this paper are concerned,investor sentiment has a high degree of correlation with SSE 50 Index.The investor sentiment has a dynamic correlation with SSE 50 Index,investor sentiment will affect SSE 50 Index,and SSE 50 Index will also react on investor sentiment.Finally,that paper puts forward reasonable suggestions to investors,listed companies and market regulators,aiming to help investors avoid risks to prevent irreparable losses.It ensures the trust of Listed Companies in the hearts of shareholders and establish a good reputation of the company,which is more conducive to the long-term development of the company.And it helps market regulators to maintain the order of the stock market,and avoid the spread of market risks.It also ensures the safety of the stock market The market has been developing well for a long time.