Research On Stock Index Prediction Based On Event Driven Model

The indicators obtained in previous studies are usually based on only one data source and thus may not fully cover the factors that can affect the stock market movements.With the development of the stock market and the rise of social networks,more and more investors have begun to discuss the topics about the stock market,and the reports of the stock market have become more abundant.With the rapid development of stock information,the analysis of the stock market has become more diversified and forecasting the stock market movements based on the internet has become one of the research hotspots.However,how to extract the events contained in news and sentiments contained in social networks,and to integrate them with the quantitative data of the stock market to predict the stock index is still a challenge.In this work,we first propose a method based on RBM to extract the events of news,which can transform a piece of news into a vector with a fixed length.Then we propose a sentiment-topic model named LDA-S to extract the sentiments of short texts.The model is based on LDA and improves the Gibbs Sampling method of LDA.At the same time,a priori information is added,and the sentiment distribution of documents can be extracted according to the topics of different documents.Next,we introduce the event extraction method based on RBM,which can extract structured information from original news and express the news as vectors.In order to improve the prediction for stock market composite index movements,we exploit the consistencies among different data sources,and develop a multi-source multiple instance model that can effectively.combine events,sentiments as well as the quantitative data into a comprehensive framework.To effectively capture the news events,we successfully apply a novel event extraction and representation method.In addition,our approach is able to automatically determine the importance of each data source and identify the crucial input information that is considered to drive the movements,making the predictions interpretable.Based on the experiments,it is verified that the multi-source multiple instance model and the feature extraction method can improve the performance of the prediction for stock market composite index movements.