基于自然最近邻居的社团检测算法
针对传统社团检测算法无法判断网络中特殊节点和SCAN算法对于参数依赖性太大的缺点,提出了一种基于自然最近邻居概念的社团检测算法CD3N。算法利用自然最近邻居无参的特性,首先以结构相似度为基准,计算出网络节点的自然最近邻居,并依此构造小值最近邻域图;然后取邻域图中邻居数最多的节点为核心节点,根据可达关系,构造关于核心节点的社团;重复选取核心节点并构造社团的过程,直到没有可归入社团的节点。将算法应用到空手道俱乐部网络和海豚网络中,并与SCAN算法进行对比。实验结果表明,CD3N算法有效解决了参数敏感性问题,能够很好地进行社团检测。...
Saved in:
| Published in | 计算机应用研究 Vol. 31; no. 12; pp. 3560 - 3563 |
|---|---|
| Main Author | |
| Format | Journal Article |
| Language | Chinese |
| Published |
重庆大学计算机学院软件理论与技术重庆市重点实验室,重庆,400044
2014
|
| Subjects | |
| Online Access | Get full text |
| ISSN | 1001-3695 |
| DOI | 10.3969/j.issn.1001-3695.2014.12.011 |
Cover
| Summary: | 针对传统社团检测算法无法判断网络中特殊节点和SCAN算法对于参数依赖性太大的缺点,提出了一种基于自然最近邻居概念的社团检测算法CD3N。算法利用自然最近邻居无参的特性,首先以结构相似度为基准,计算出网络节点的自然最近邻居,并依此构造小值最近邻域图;然后取邻域图中邻居数最多的节点为核心节点,根据可达关系,构造关于核心节点的社团;重复选取核心节点并构造社团的过程,直到没有可归入社团的节点。将算法应用到空手道俱乐部网络和海豚网络中,并与SCAN算法进行对比。实验结果表明,CD3N算法有效解决了参数敏感性问题,能够很好地进行社团检测。 |
|---|---|
| Bibliography: | ZHU Qing-sheng,JIANG Tian-hong,ZHOU Ming-qiang（Chongqing Key Laboratory of Software Theory ＆ Technology, College of Computer Science, Chongqing University, Chongqing 400044, China ） 51-1196/TP community detection;complex network;natural nearest neighbors;structural similarity To void the problem that the traditional detection algorithm can’t determine special nodes in the network and the SCAN algorithm’s parameter sensitivity,this paper proposed a community detection algorithm based on natural nearest neighbor（ CD3N）. According to structure similarity,it calculated natural nearest neighbor of each nodes,constructed neighborhood graph. Then it chose node which had maximum number of neighbors as core node,constructed community on this core node. It repeated the process to select a core node and construct community until no nodes left,and applied this algorithm in the Zachary’s Karate Club network and Dolphin network,and compared with SCAN algorithm. The results of experiment show that the CD3 N algorithm is able t |
| ISSN: | 1001-3695 |
| DOI: | 10.3969/j.issn.1001-3695.2014.12.011 |