Research On The Interactive Relationship Between Media Attention,Media Sentiment And Stock Market Under Big Data

With the continuous opening of China's financial market,financial media reports have gradually highlighted the pricing function of capital.Financial news is the main medium for investors to obtain macroeconomic financial information and collect dynamic news of listed companies.Independent thirdparty financial media in China's capital market affects stock market performance while serving investors.Financial media reports and reports in the era of mobile internet the linkage relationship between stock price and price has become a research hotspot in the field of finance.At the same time,big data and artificial intelligence technologies in the field of computer science also provide favorable conditions for the automated and intelligent mining and analysis of massive financial information,especially the maturity of Chinese natural language processing technology,making behavioral finance Scholars in the field can more systematically and meticulously study the relationship between financial media reports and asset pricing.Based on this,this paper firstly uses the web crawler technology to obtain more than 1,600,000 financial news reports from the listed companies of the A-share CSI 300 Index from the beginning of 2014 to the end of 2017,and use the support from the mainstream financial websites and media in China.The machine learning classification method of vector machine and the text sentiment analysis technology based on text sentiment dictionary are used to classify statistics and sentiment analysis.At the same time,this paper builds weekly "media attention" and "based on two factors: news quantity and news sentiment tendency".Agent indicators of media sentiment,using dynamic panel stepwise regression model classification to explore the linkage relationship between financial news reports and stock price in China.Through the research,this paper finds that under the environment of China's weak and effective capital market,the financial media's news reports are generally optimistic.The media attention and media sentiment triggered by media reports can stimulate market trading enthusiasm and form a certain market price.Pressure,media attention has a greater impact on stock trading volume,and media sentiment has a greater impact on stock prices.Among them,positive media reports can stimulate market enthusiasm to promote stock price increase,while negative media reports are more likely to cause large deviations in stock prices and longer impact on subsequent market transactions;media during trading hours and non-trading periods The report will stimulate the trading sentiment of the market,but the news released during the trading session has a greater impact on the stock trading volume,while the news released during the non-trading period has a greater impact on the opening price of the first trading day.After a comparative analysis,this paper finds that the media reports in the bull market are more optimistic than the bear market.The media reports are more likely to stimulate the market to trade and promote the stock price rise,and the market is more sensitive to positive news reports.Based on the release properties of financial news,this paper uses the financial news text big data to innovatively construct the classification agent indicators of “media attention” and “media sentiment”,which provides new ideas for portraying the influence of media reports.Moreover,this article also The two perspectives of "media attention" and "media sentiment" systematically and comprehensively studied the linkage relationship between mainstream financial media reports and stock price and price in China,and provided research for "media reporting and asset pricing" in the field of behavioral finance.New perspectives and evidence also provide a new direction for investors to more effectively mine financial news information to assist investment decisions.