首页 | 本学科首页   官方微博 | 高级检索  
     检索      

一种优化共同邻居影响的动态距离社区发现算法
引用本文:万甲鑫.一种优化共同邻居影响的动态距离社区发现算法[J].教育技术导刊,2009,19(10):142-145.
作者姓名:万甲鑫
作者单位:华东师范大学 通信与电子工程学院,上海 200241
摘    要:在众多社区发现算法中,Attractor算法是一种快速的社区发现算法,具有社区检测准确率高的优点。为解决Attractor算法在距离更新过程中节点对度值相差太大,影响小度节点所属社区判断问题,提出一种优化共同邻居影响的Attractor社区发现算法。该算法在Attractor算法提出的动态距离节点交互模型基础上,考虑节点对两者度值差异,通过在节点对与共同邻居交互模式中增加一个大度节点不利系数,以增加小度节点对邻居的吸引作用。采用LFR基准网络,在不同结构网络上验证改进算法的有效性。实验结果表明,改进算法与Attractor算法相比社区发现准确度更高。

关 键 词:社区发现  复杂网络  动态演化  动态距离  
收稿时间:2020-04-09

A Community Detection Algorithm by Optimizing the Influence of Common Neighbors Based on Distance Dynamics
WAN Jia-xin.A Community Detection Algorithm by Optimizing the Influence of Common Neighbors Based on Distance Dynamics[J].Introduction of Educational Technology,2009,19(10):142-145.
Authors:WAN Jia-xin
Institution:School of Communication and Electronic Engineering, East China Normal University, Shanghai 200241, China
Abstract:Among many community discovery algorithms, the attractor algorithm is a fast community discovery algorithm, and has the advantages of high community detection accuracy. In order to solve the problem that the difference of degree value between nodes in the process of distance update affects the community judgment of small degree nodes, a community discovery algorithm of attractor which optimizes the influence of common neighbors is proposed. Based on the dynamic distance node interaction model proposed by the attractor algorithm, this algorithm takes into account the difference of the degree values between the two nodes. By adding a negative coefficient of large nodes in the interaction mode between the node pair and the common neighbor, the attraction of the small node to the neighbor is increased, and the LFR benchmark network is used to verify the effectiveness of the improved algorithm on different networks.
Keywords:community detection  complex network  dynamic evolution  distance dynamics  
点击此处可从《教育技术导刊》浏览原始摘要信息
点击此处可从《教育技术导刊》下载免费的PDF全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号