【摘要】 网络中的社区发现是当前的一个研究热点。在众多社区发现算法中,标签传播算法因简单快速而被广泛应用。但标签传播算法也存在结果稳定性较差的问题。通过提取非重叠完全子图来避免社区重叠,提取最小极大团来避免巨型社区的出现,基于此,对标签传播算法的初始化过程进行改进,提出了一种稳定的标签传播社区发现算法,即非重叠最小极大团提取算法。在真实网络中的实验结果表明该算法可以大幅提高结果的稳定性。更多还原

【Abstract】 Community detection in networks is a hot research topic currently.Among many community detection algorithms,label propagation algorithm is widely used for it is simple and rapid.But label propagation algorithm also has the problem of poor stability results.To avoid overlapping community by extracting non-overlapping complete subgraph,and to avoid huge community by extracting minimal maximal clique,the initialization process of label propagation algorithm was improved to derivea stable label propagation algorithm,namely disjoint minimal maximal clique algorithm.Experimental results on real-world networks show that the new method improved the stability greatly.更多还原