Abstract: With the growing of the Internet,the network of large data has become an important distribution center for the financial sector.These data give valuable opportunities and severe challenges to stock analysis and prediction.The development of Web2.0 allows investors to actively participate in all aspects of the creation of network of information,communication and accessing.This paper disclosed information on the Internet and generated complete and effective fundamental of information to get 10 actors that may impact the stock market,which are economic and cycle,fiscal policy,changes in interest rates and exchange rates,price changes,inflation,political policies,changes in the industry,operating conditions and the downstream effects.We proposed an algorithm for the stock market prediction based on these 10 factors.Experiments show that the method has good feasibility,and can bring significant academic and commercial values.

Key words: Data mining,Stock price forecast,Web big data