Optimized Stock Price Prediction Based On Investor Sentiment And Stock Recommendation Display Board Research

With the development of technology and the popularization of the Internet,social media with forums,microblogs,news,and other community communication sites as the main carrier is rapidly enriching people's leisure lives.In these Internet media,everyone can publish Own opinions,publish information in the form of publishing opinions,and at the same time publish opinions,also disseminate the opinions of others,and also be influenced by the opinions of others,that is,the audience of media information.Under the influence and guidance of these social media,people's investment methods are gradually changing,and the Internet is the main carrier of social media.As a financial product that is well known to most people,the price fluctuations of stocks can affect many aspects of people's daily lives in social and economic life.Therefore,some methods and models are used to predict the future changes in stock prices.Great research value.The rise and fall of stock prices are affected by many factors,and there is a memory relationship between time series data of stock prices.The Long Short-Term Memory networks model is an improved model of the RNN neural network model,which can more effectively learn the rules in long-term dependencies.Therefore,it is dealing with time series data with long-term dependencies such as stock data More advantages.Therefore,based on the sentiment analysis of the stock review text,this paper builds a stock price trend prediction model based on long-term and short-term memory neural network(LSTM).Based on the traditional short-term prediction model of LSTM stock price based on investor sentiment,this article takes the improvement of forecasting accuracy as an example.In light of the previous classic model,it ignores the influence of investor sentiment changes on stock market changes.Attention indicators that reflect the attention of investors and sentiment indicators that can reflect investors' investment sentiment can enhance the performance of prediction models.With the help of investor sentiment indicators and investor attention indicators,we can more accurately analyze the emotional tendencies and behavior characteristics of investors in China's stock market,and then further analyze the impact of investor sentiment on stock prices.Therefore,this paper takes the traditional financial time series model as the reference model,and compares the LSTM model with the traditional financial time series prediction model,which proves that the LSTM model has significantly improved the accuracy of prediction.Then,for the above-mentioned short-term stock price prediction model,a concise and easy-to-understand stock recommendation board is constructed for non-professional investors,which is mainly based on the investor's risk appetite,investment volume,or type of stocks of interest.The relevant information of the stock is displayed to provide reference and suggestions for the investment behavior of stock investment users.