基于文本挖掘的网络信息与股票市场关联机制研究

项目名称： 基于文本挖掘的网络信息与股票市场关联机制研究

项目编号： No.71303210

项目类型： 青年科学基金项目

立项/批准年度： 2014

项目学科： 管理科学

项目作者： 杨晓兰

作者单位： 浙江大学

项目金额： 21万元

中文摘要： 在互联网时代，提取网络信息中体现的投资者情绪特征，研究网络信息与股票市场的关联机制具有重要的理论意义与现实意义。由于中国股票市场上个人投资者占主体地位，且短线投资者众多，因而对即时信息的依赖程度较高，这无疑给网络信息引起股市变动造就了更大空间。课题立足中国股票市场实际，综合运用多学科方法，探究我国股票市场网络信息影响股票市场的途径和机理。研究主要围绕以下方面展开：网络环境下股票市场信息生成与传播的一般过程、特征及对市场参与者行为模式的影响；网络信息引发社会互动以及投资者情绪传染、放大，并影响股票市场的机理；运用文本挖掘技术，构建网络情绪指标,实证检验网络信息情绪与股票市场的关联机制；结合股市价格变化极端情况的典型案例，探索网络信息引发群体情绪极化的条件，为投资者决策和完善网络信息监管、维护股市稳定提供政策建议。

中文关键词： 网络信息；股票市场；投资者情绪；文本挖掘；

英文摘要： In the era of Internet, it is meaningful and important to analyze the effect of Internet information on stock market by abstract the properties of investment sentiment involed in the Internet information. Since Chinese stock market has numerous individual investors who are highly interested with the Internet information, stock market movements would be caused by the Internet information greatly. Based on the phenomena of Chinese stock market, this project utilizes multi-disciplinary approaches to explore the relationship between Internet stock market information and the movement of stock market. This study mainly focuses on the following issues: Firstly, under the Internet environment, what are the general process, and features of the generation and diffusion of stock market information, and how dose stock market information influence investors behavior. Secondly, how does the Internet information affect social interaction,sentiment contagion among investers? Thirdly, we will use text mining method to build a Internet sentiment indicator and empirically test the linkage between Internet information sentiment and the stock market. Finally, based on the typical cases of extreme stock price fluctuations, we will explore the conditions of emotional polarization evoked by Internet information and provide some suggest

英文关键词： Internet information；Stock market；Investor sentiment；Text mining；