基于核心节点扩散的有向网络重叠社团发现算法

复杂网络中的社团结构探测是当前复杂网络研究领域的一个热点问题。传统的社团划分算法主要以无向、无权网络作为分析对象,不能够适用于现实世界中各种有向网络、加权网络。在分析和研究各种社团划分算法的基础上,提出一种新的重叠社团发现算法。该算法从网络中的核心节点开始,不断合并适应度最大邻居节点,最终将网络划分为多个重叠的社团。最后,将该算法应用到两个有向网络中,实验表明该算法能够很好地划分出有向网络中的重叠社团。

Algorithm of Overlapping Community Detection in Directed Networks Based on Key Node Spreading

The problem of detecting community structure in large-scale directed networks is increasingly attracting research attention. The previous research mainly focuses on detecting community structure in undirected network, they do not work when the network is directed and weighted in the real world. Based on the analysis and research on the basis of various community detecting algorithm, proposes a new al- gorithm to detect overlapping communities. The algorithm starts from the core nodes in the network, and constantly merge fitness maxi- mum neighbor nodes, finally divides the network into multiple overlapping community. Tests the algorithm on two directed network, exper- imental results show that the algorithm is rather efficient to detect overlapping communities of directed network.