计算机工程

• 专栏 • 上一篇    下一篇

一种联合拓扑与属性的社区模糊划分算法

郭进时,汤红波,葛国栋   

  1. (国家数字交换系统工程技术研究中心,郑州 450002)
  • 收稿日期:2012-10-18 出版日期:2013-11-15 发布日期:2013-11-13
  • 作者简介:郭进时(1987-),女,硕士研究生,主研方向:移动网络,社会网络;汤红波,教授;葛国栋,博士研究生
  • 基金项目:
    国家“863”计划基金资助项目(2011AA7116031, 2011AA010604)

Community Fuzzy Dividing Algorithm Combining Topology and Attribute

GUO Jin-shi, TANG Hong-bo, GE Guo-dong   

  1. (National Digital Switching System Engineering & Technological R&D Center, Zhengzhou 450002, China)
  • Received:2012-10-18 Online:2013-11-15 Published:2013-11-13

摘要: 现有的社区发现算法通常基于结构特性进行社区划分,对节点属性特征欠缺考虑。为此,提出一种基于模糊等价关系的社区发现算法。用完全相异距离指数的概念将拓扑结构与属性特征相结合,以此作为隶属关系建立模糊等价关系矩阵,选择合适的聚类阈值对网络进行社区划分。实验结果证明,与传统的GN算法相比,该算法发现社区的准确率较高,在相同社区内的节点连接紧密且具有同质性。

关键词: 社会网络, 社区发现, 属性, 完全相异距, 模糊矩阵, 等价关系

Abstract: Aiming at the problem that most existing community detecting algorithms are usually based on the structure characteristics of network and lack of consideration attribute information, a community detection algorithm is proposed based on fuzzy equivalence relation combining topology and attribute in social networks. In this algorithm, a new concept of integrated dissimilarity distance index is used for combining topology and attribute, and it is regarded as the subordinate relation to build the fuzzy equivalence relation matrix, appropriate clustering threshold value is choses for community detection. Experimental result proves that the algorithm has high accuracy compared with those traditional GN algorithms, and nodes in the same community are densely connected as well as homogeneous.

Key words: social network, community detection, attribute, complete dissimilarity distance, fuzzy matrix, equivalence relation

中图分类号: