【摘要】 由于能够反映网络内部结构,重叠社区划分在各领域有着越来越重要的作用。LFM算法是其中较为流行的一种社区划分方法。但其存在一些缺点,例如在网络变得庞大和复杂的时候,时间消耗会变得巨大。为了解决这一问题,提出核心区域的概念,并藉此对LMF算法进行改进。最后通过实验验证,发现该算法能够减小时间消耗,同时能够得到更为可靠的社区划分。更多还原

【Abstract】 Overlapping community detection has become more and more important since it can reveals the inner structure of networks. LFM algo-rithm is one of the most popular way to detect communities in complex networks, however the algorithm itself has some weaknesses, such as large time consumption when the network become large and complex. To overcome these problems, based on LFM, presents an im-proved LFM algorithm(M-LFM) which proposes a definition of core area and apply it into the process of community detection with LFM.Experiments on real networks and artificial networks show that the improved algorithm can decrease time consumption and get better result than LFM.更多还原