一种基于相对关系亲密度的局部社团发现算法

A Local Community Detection Algorithm Basedon Relative Intimacy

在复杂网络研究中,社会网络中的社团发现,对商业营销、疾病传播控制等具有实际意义。目前许多研究针对于全局网络进行社团挖掘。挖掘算法因其较高的复杂度往往不适用于动态、大型网络。针对某个点或者某个区域的局部社团挖掘成为了近期的研究热点。为此提出了一种新的相对关系亲密度的计算方法,并与已有算法思想结合,形成了一种新的局部社团发现算法,提高了算法性能。基于已知社会网络、随机测试网络的实验证明了算法的有效性。

In the study of complex networks, social network community detection has practical significance such as commercial marketing, disease control, etc. Many studies are conducted for the detection in the whole network, which is often not available for dynamic, large-scale networks. Therefore, local community detection has become a hot topic recently. This paper presented a new method for calculating the relative intimacy between nodes, and combined with the existing algorithms thought, forming a new local communities detection algorithm to improve the performance. Experiments based on the known social networks and random testing network proved the effectiveness of the algorithm.