信息网络的社区发现及其应用研究

Studies on Community Detection and Its Application in Information Network

在对信息网络社区发现研究的基本概念进行简单介绍的基础上,将目前的典型算法初步划分为5类,即传统发现方法、基于分割的方法、基于模块性质量优化的方法、基于动态模型的方法与基于谱分析的方法,然后着重对各种发现方法进行分类分析与比较,并以Web信息网络为例简述社区发现的应用,最后指出了目前方法存在的缺陷与未来可能的改进方向。

Based on the introduction of concepts and principles related with the study,we classify existing popular algorithms into 5 categories： traditional algorithms,partition based,modularity optimization based,dynamic model based and spectral based algorithms.Comparison and analysis of various detection methods are emphasized,on this base,we illustrate the applications of community detection with web network,and finally we point out the defects of the current methods and the possible future directions to improve.