力导向模型的复杂网络社区挖掘算法

在复杂网络中发现和刻画社区结构是近年来复杂网络研究的重点方向之一。提出了一种社区挖掘的新思路,即根据力导向模型的原理,通过计算社区与节点之间的作用力来决定节点的社区归属。根据该思路设计了基于力导向模型的算法框架FDCD(Force-directed Community Detect),并利用FR模型、KK模型、LL模型和Q模型进行了验证。实验表明,基于FDCD算法框架的多种不同算法不仅能较好地识别社区结构,而且基于LL模型的FDCD算法达到了线性计算复杂度,能适用于大规模网络的社区挖掘。

A community detect algorithm on force-directed model

The problem of detecting and characterizing community structure is one of the outstanding issues in the study of complex network. This paper propose a new community detect algorithm base on Force-directed model, which categorize the point to community decided by the force between them. This paper design an algorithm named Force-Directed Community Detect, FDCD, and implement algorithm using four different Force-directed Model. The experiments show that four algorithms can find community in real social network with high Q Modularity. Especially, each efficiency of algorithm base on LL model is o(n), which is fit for community detect in large network.