基于模块化多尺度算法的通信重叠社区检测  

Overlapping Community Detection Algorithm in Communication Based on Modular Multi-scale

作　　者：蒋智恩 JIANG Zhi-en(Department of business administration,Fujian Polytechnic of Information Technology,Fuzhou 350003,China)

机构地区：[1]福建信息职业技术学院商贸管理系,福建福州350003

出　　处：《信阳农林学院学报》2020年第2期121-126,共6页Journal of Xinyang Agriculture and Forestry University

摘　　要：针对常规的社区检测方法未能量化社区特征这一问题,通过考察社区结构的多尺度特征,提出一种基于模块化多尺度的重叠社区检测算法。从节点、社区和网络三个尺度分别量化社区结构,并建立高度模块化的社区检测模型。首先提取节点集,扩展社区形成社区初始集,最后通过算法迭代聚类特定节点。在人工网络和现实网络中对算法进行了大量测试,几种质量评价指标以及实验结果均表明该算法计算成本低,且准确高效。Previous methods of community detection cannot quantify community features.To address the problem,by examining the multi-scale features of community structure,an overlapping community detection algorithm was proposed based on modular multi-scale.Community structure was quantified in terms of features for individual node,individual community and the whole network,and a highly modular community detection model was developed.Firstly,nodes sets were extracted,and the community was expanded to form community initial set.Specific nodes were clustered by iterative algorithm finally.Extensive tests were carried out on artificial network and real network.The results show that the algorithm has low computational cost and efficient and accurate performance.

关 键 词：重叠社区检测 模块化 多尺度特征 节点度 连接密度 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]