网络话题活性模型的仿真与分析

Simulation and Analysis of Internet Topic Active Model

热点话题发现是互联网舆论研究中的基础性问题之一,为此提出了话题活性模型和快速筛选算法。话题活性模型能够在话题形成讨论热点的早期将其从混合的帖子到达流中分离出来;快速筛选算法仅根据即时信息进行计算,不需要长期历史信息,且计算的时间复杂度仅为O(n),较好的解决了热点话题快速筛选问题。最后仿真和实证,证明了该方法的有效性。

Finding hot topic is a basic issue of internet public opinions research, in order to resolve this problem, topic active model and fast filter hot topics on BBS （electronic bulletin board system） arithmetic were advanced. Topic active model can find out hot topics from mixed bulletins during their discussion forepart, and fast filter just needs current information not history data and can calculate during O （ n ） time. The result of simulation shows that this topic activity model is effective.