大数据下媒体关注度与股票收益关系研究  

Relationship Between Media Attention and Stock Return Based on Big Data

作　　者：王春峰[1] 李嘉毅[1] 房振明[1] 

机构地区：[1]天津大学管理与经济学部,天津300072

出　　处：《天津大学学报（社会科学版）》2016年第2期103-108,共6页Journal of Tianjin University：Social Sciences

基　　金：国家自然科学基金资助项目(71271146)

摘　　要：互联网的发展使得市场信息更容易获得,互联网已经成为投资者获取有价值信息的主要途径,也对投资者行为逐渐产生影响。通过自主开发的金融数据挖掘系统,采集了15个网站150多个与股票相关栏目的 100余万条媒体报道信息,并能够实现对信息更精细的统计,使得媒体关注度能够在日内按照不同时间段进行细分。在此基础上,研究交易时间和非交易时间媒体关注度与股票收益之间的关系。结果表明,交易时间的媒体信息对部分股票收益率产生显著影响,非交易时间媒体信息影响较小,虽然这种情况目前存在于部分股票中,但随着互联网进一步发展,网络信息将逐步成为需要考虑的风险因素。Thanks to the development of Internet,market information can be easier to obtain,and investors can get valuable information from it. Thus Internet gradually affects the investors＇ behavior. One million media reports which are collected by financial data mining system including 15 websites and 150 columns about stocks could be accurately calculated.In this way,media attention can be subdivided according to different period of time in a day. Based on these researches,the author analyses the relationship between media attention and stock return during trading and no-trading hours. The result shows that media information could significantly affect the return for some stocks during trading hours,while less affect during no-trading hours. Internet information will be the important risk factor with the development of internet,though the state has only occurred in some stocks.

关 键 词：数据挖掘 媒体关注度 互联网信息 百度指数 

分 类 号：F830[经济管理—金融学]