突发重大新闻事件中基于兴趣的网民活跃度模型

INTEREST-BASED USER ACTIVITY MODEL IN BREAKING NEWS EVENTS

从系统科学的视角出发,将受到外界媒体报道激励并作出某种可记录的活跃行为的网民群体视作一个智能系统,提出了突发重大新闻事件中基于兴趣的网民活跃度模型.该模型将外界媒体的激励、网民之间的相互激励以及网民自身的遗忘作为影响网民活跃度的三个重要因素.从单个网民的活跃行为入手,推导出整个网民群体活跃度的数学表达式.然后以搜索行为为例,对该模型的有效性进行了验证.并将该模型与ETH发布的SIR,模型和NetLogo提供的流言工厂模型进行对比实验,在此基础上,阐述了所提模型的不同之处.最后讨论了该模型的局限性和未来的工作.

From the perspective of the systems science, a group of Internet users who are inspired by a breaking major news event and perform a recordable behavior, can be treated as an intelligent system. An interest-based user activity model is thus proposed in this paper. Three significant factors which can shape the users＇ behavior are taken into consideration： The news media as an outside stimulus, the interactions between users, and the natural fading memory of the users. The modeling work begins with the behavior dynamics of an individual to derive the mathematical model of the entire group of Internet users. The model is verified by empirical data sets from the search engine, Baidu. Then it is compared with the SIR model published by ETH, and the rumor mill model provided by NetLogo. Finally, the limitation of the model is described.