基于极大团的大规模网络社区发现算法

摘要

In order to discover the community structure in large-scale networks,a community mining algorithm based on maximum clique is proposed.By introducing an idea of maximum clique,and via locating the key clique structure regarded as the initial community by with key nodes,it is determined by quantifying the similarity between the neighbor node and the community whether the neighbor node is merged into the community,thus to acquire a more reasonable community structure.And comparison of the algorithm with the representative CPM algorithm in four real networks indicates that this algorithn could find the community structure of large-scale network and that the rationality of this community structure is better than that by CPM algorithm.This algorithm described in this paper is an effective community discovery algorithm of social network.%为了能够发现大规模网络中的社区结构,提出一种基于极大团发现的社区挖掘算法.引入极大团思想,通过关键节点定位关键团结构作为初始社区,并通过量化初始社区的邻居节点与该社区的相似度,判断该邻居节点是否归并到该社区,从而得到较为合理的社区结构.将改进算法与具有代表性的CPM算法在四个真实网络上进行对比实验,实验结果表明,改进算法不仅可以发现大规模网络的社区结构,而且所发现的社区结构合理性优于CPM算法,是一种有效的社会网络社区发现算法.