Stock Trend Prediction by Classifying Aggregative Web Topic-Opinion

摘要

According to the Efficient Market Hypothesis(EMH) theory, the stock market is driven mainly by overall information instead of individual event. Furthermore, the information about hot topics is believed to have more impact on stork market than that about ordinary events. Inspired by these ideas, we propose a novel stock market trend prediction method by Classifying Aggregative Web Topic-Opinion(CAWTO), which predicts stocks movement trend according to the aggregative opinions on hot topics mentioned by financial corpus on the web. Several groups of experiments were carried out using the data of Shanghai Stock Exchange Composite Index(SHCOMP) and 287,686 financial articles released on SinaFinance, which prove the effectiveness of our proposed method.