Complete graph model for community detection
Complete graph model for community detection
复制标题
DOI：
10.1016/j.physa.2016.12.014
复制
发表时间：
2017-04
影响因子：
3.3
通讯作者：
P. Sun;Xiya Sun
中科院分区：
文献类型：
--
作者：
P. Sun;Xiya Sun
社区检测带来了许多问题，多年来一直受到人们的关注。本文提出了一个新的框架，试图在同一个度量的基础上，完整图模型的社区内部和外部的测量。特别是，外部被建模为一个完整的二分。我们通过最大化子网络内部和外部之间的差异来将网络划分为子网络。此外，我们将我们的方法与基于LFR基准的计算机生成网络以及真实网络上的一些最先进的方法进行比较。实验结果表明，该方法在社区检测方面取得了较好的效果，能够对不规则网络进行分裂，在空手道网络和海豚网络上取得了较好的效果。
Community detection brings plenty of considerable problems, which has attracted more attention for many years. This paper develops a new framework, which tries to measure the interior and the exterior of a community based on a same metric,completegraph model. In particular, the exterior is modeled as a complete bipartite. We partition a network into subnetworks by maximizing the difference between the interior and the exterior of the subnetworks. In addition, we compare our approach with some state of the art methods on computer-generated networks based on the LFR benchmark as well as real-world networks. The experimental results indicate that our approach obtains better results for community detection, is capable of splitting irregular networks and achieves perfect results on the karate network and the dolphin network.