基于GARCH模型的网络新闻与舆情的波动性分析

摘要

网络新闻产生的舆情波动一般具有异方差特征,难以用普通模型拟合.由诺贝尔经济学奖获得者恩格尔教授提出的条件异方差(CARCH)模型在分析证券价格波动性方面获得极大成功.本文利用GARCH模型分析网络新闻与舆情的波动性,通过典型事件的舆情采集,分析数据的特征.研究表明,网络新闻与舆情的波动性符合GARCH模型的特征,通过参数的调整和检验,可以实现模型与数据的良好拟合.%The volatility of public opinion caused by Web news has the character with heteroskedasticiry and is not to be fit by common models easily. GARCH (Generalized Auto-Regressive Conditional Heteroskedasticiry) model proposed by professor En-gle is success to analyze the volatility of stock price. This paper uses GARCH model to analyze the volatility of Web news events and public opinions by the data coming from typical news events in famous Web. The result shows that the volatility of Web news events and public opinions is suitable to GARCH model by adjusting and testing of parameters, and has a good fitness.