[1]辛慧英,刘向阳.基于核心度和偏移量的社区检测算法[J].计算机技术与发展,2020,30(10):37-41.[doi:10. 3969 / j. issn. 1673-629X. 2020. 10. 007]
　XIN Hui-ying,LIU Xiang-yang.Community Detection Algorithm Based on Core Degree and Distance[J].,2020,30(10):37-41.[doi:10. 3969 / j. issn. 1673-629X. 2020. 10. 007]
点击复制

基于核心度和偏移量的社区检测算法()
分享到：

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:
30
期数:
2020年10期
页码:
37-41
栏目:
智能、算法、系统工程
出版日期:
2020-10-10

文章信息/Info

Title:
Community Detection Algorithm Based on Core Degree and Distance
文章编号:
1673-629X(2020)10-0037-05
作者:
辛慧英; 刘向阳
河海大学 理学院,江苏 南京 211100
Author(s):
XIN Hui-ying; LIU Xiang-yang
School of Science,Hohai University,Nanjing 211100,China
关键词:
边介数; 距离矩阵; 核心度; 偏移量; 核心社区
Keywords:
betweenness; distance matrix; core degree; distance; core community
分类号:
TP301. 6
DOI:
10. 3969 / j. issn. 1673-629X. 2020. 10. 007
摘要:
为减少社区检测算法中大量中间结果的计算对社区划分的影响,同时能够准确检测到网络的社区划分以及网络的核心社区,提出了一种基于核心度和偏移量的社区检测算法,其中核心度和偏移量定义了任意节点作为社区核心的程度。 首先针对复杂网络的邻接矩阵,应用广度优先搜索算法计算网络中节点之间的边介数,基于边介数确定网络中每条边的权值,计算得到网络的加权邻接矩阵及全局距离矩阵;然后计算网络节点的核心度和偏移量,来确定社区的核心节点和核心社区;最后对其余节点进行划分以完成社区检测。 在数据集 Karate、Dolphins、Football 上的实验结果表明,该算法具有很好的稳定性,并且可以很好地检测出社区结构,相比其他的方法,该算法复杂度更低,计算量更少,更高效。
Abstract:
In order to reduce the impact of the calculation of a large number of intermediate results in the community detection algorithms on community partitioning,and to accurately detect the community division of the network and the core community of the network,we propose a community detection algorithm based on core degree and distance which define the degree to which any node is the core of the community. Firstly,based? ? ? ?on the adjacency matrix of the complex network,the breadth-first search algorithm is applied to calculate the betweenness in the network. The weight? of each edge is determined based on betweenness,and the weighted adjacency matrix and global distance matrix of the network are calculated. Then,the core degree and distance of the network node are calculated to determine the core nodes and core communities. Finally,the remaining nodes are dispatched to complete the community detection. The experimental results on the datasets Karate, Dolphins, and Football show that the proposed algorithm can well detect the community structure with high stability. Compared with other methods,it has lower complexity, less calculation and more efficiency.

相似文献/References:

[1]赵礼峰,黄奕雯. 基于矩阵运算K短路径算法[J].计算机技术与发展,2017,27(04):98.
　ZHAO Li-feng,HUANG Yi-wen. Research on K-shortest Path Algorithm with Matrix Operations[J].,2017,27(10):98.
[2]华 超,刘向阳.基于密度加权原型网络的小样本学习算法[J].计算机技术与发展,2022,32(09):8.[doi:10. 3969 / j. issn. 1673-629X. 2022. 09. 002]
　HUA Chao,LIU Xiang-yang.Few-shot Learning Based on Density-weighted Prototypical Network[J].,2022,32(10):8.[doi:10. 3969 / j. issn. 1673-629X. 2022. 09. 002]

更新日期/Last Update: 2020-10-10