A Hierarchical Attention Network For Stock Prediction Based On Attentive Multi-view News Learning

Stock prediction refers to the prediction behavior of the future stock trend direction and the degree of fluctuation based on the development of the stock market by researchers who have a deep understanding of the stock.However,due to the high volatility and non-stationarity of the stock market,accurate stock prediction is very difficult.The explosive growth of news media information and the continuous development of natural language processing and text mining technology provide new ideas for further research on stock forecasting,enabling researchers to reveal market trends and volatility from numerous news media information.Among the existing stock prediction methods based on news information,most of them generally based on single news view,e.g.,headline or body,as a predictive indicator and thus information received is insufficient or incomplete which also lacks of study on market information,then bring low performances of models.In this research,we propose a hierarchical attention network based on attentive multi-view news learning(NMNL)to excavate more useful information from news and the stock market for stock prediction.The core of our approach is a news encoder and a market information encoder.In the news encoder,we learn multi-view news information representation from news headlines,bodies and sentiments by regarding them as three independent parts.We find that the combination of headline,body and sentiment outperforms conventional models on single news view.In the market information encoder,we employ the attention mechanism to capture pivotal news information and combine technical indicators to represent representative market information.In addition,a temporal auxiliary based on Bi-directional Long Short-Term Memory(Bi-LSTM)model is used to generate the contextual market information for stock prediction.Extensive experiments demonstrate the superiority of NMNL,which outperforms state-of-the-art stock prediction solutions with an average Directional Accuracy(DA)of 0.608 and Matthews Correlation Coefficient(MCC)of 0.1072 on HS 300,respectively.The hierarchical attention network for stock prediction based on attentive multi-view news learning proposed in this thesis can effectively dig out rich news information from different views from a large number of financial news,so as to enhance the representation ability of news information;It can learn representative market information from different information sources of news information and fundamentals,and improve the accuracy of stock prediction.It provides a new idea for how to analyze valuable financial text information and enhance the representation ability of market information to improve the accuracy of stock prediction.