• 中国计算机学会会刊
  • 中国科技核心期刊
  • 中文核心期刊

J4 ›› 2016, Vol. 38 ›› Issue (02): 363-369.

• 论文 • 上一篇    下一篇

一种基于聚集系数的社区发现算法

樊梦佳1,钮艳2,杜翠兰2,张仰森1   

  1. (1.北京信息科技大学智能信息处理研究所,北京 100192;2.国家计算机网络应急技术处理协调中心,北京 100190)
  • 收稿日期:2015-10-16 修回日期:2015-12-13 出版日期:2016-02-25 发布日期:2016-02-25
  • 基金资助:

    国家自然科学基金(61370139);北京市属高等学校创新团队建设与教师职业发展计划(IDHT20130519);北京市教委专项(PXM2013_014224_000042,PXM2014_014224_000067)

A community detection method based on clustering coefficient       

FAN Mengjia1,NIU Yan2,DU Cuilan2,ZHANG Yangsen1   

  1. (1.Institute of Intelligent Information Processing,Beijing Information Science and Technology University,Beijing 100192;
    (2.National Computer Network Emergency Response Technical Team/Coordination Center of China,Beijing 100190,China)
  • Received:2015-10-16 Revised:2015-12-13 Online:2016-02-25 Published:2016-02-25

摘要:

社区划分一直是复杂网络研究中的一个热门话题,社区的快速准确划分为研究复杂网络的性质提供了良好的基础。传统的社区发现方法都是在全局复杂网络的基础上进行社区划分,随着网络中节点的增加,网络规模的变大,社区发现变得更为复杂。提出了一种局部社区发现算法,该算法无需知道整个复杂网络的全部信息,只需从一个待求节点出发,考察其与邻接节点的紧密程度,逐步将邻接点添加到社区中,得到该节点所在的社区结构。同时,该算法还可实现全局网络的社区发现。利用该算法分别对Zachary空手道俱乐部网络和海豚社会网络进行社区发现,实验结果表明了该算法的准确性与可行性。

关键词: 局部社区, 社区发现, 聚集系数

Abstract:

Community detection has been a hot topic in complex network research. Fast and accurate community detection can provide a good foundation for studying the properties of complex networks. Traditional community detection methods are mainly based on the global community. However, as the number of nodes in the network increases, the network size becomes larger, so the community detection becomes more complex. We propose a local community detection method, which does not need to know the whole complex network information. Rather it just starts from an initial node and calculates the tightness between the initial node and the adjacent nodes. Adjacent nodes are gradually added to the community, and finally the community structure of the initial node is obtained. Meanwhile, this method can detect the global network community. We apply the proposed method to the Zachary karate club network and dolphin social network, and experimental results demonstrate its accuracy and feasibility.       

Key words: local community;community detection;clustering coefficient