网络舆论倾向性的动态跟踪方法及实证研究

从统计学视角审视网络舆论倾向性的监测问题,提出了以粗糙分类器为基础建立舆论倾向性分类模型,将复杂的预警指标体系简化为单一直观的跟踪统计量,并通过跟踪统计量动态跟踪舆论倾向性变化轨迹的研究方法。实证研究以2011年"郭美美事件"相关新闻跟帖为对象。分析表明,网络舆论的消极倾向性在整个事件发展过程中呈持续增长,与基本事实相符,证实了方法的可行性和适用性。

A Dynamic Method for Tracking the Tendencies of Internet Public Opinions and Its Empirical Study

This paper deals with the problems on monitoring the tendencies of Internet public opinions from the statistical view.By building the classify model based on the coarse classifier,simplifying the complex warning index system into a single and intuitive tracking statistics,the paper proposes a novel method to track the development locus of the Internet public opinions tendencies.The news threads on Jocie Kok event in 2011 are the object of our empirical study.The researching shows that the negative tendencies keep growing during the entire development process.According with the facts suggest the feasibility and applicability of this method.