摘要
针对基于中心节点的重叠社区识别算法具有发现社区不稳定的缺点,提出了一种基于中心节点和社区吸引力的重叠社区识别算法:通过节点度数以及节点与相邻节点的区域内部连通性给出中心性系数,并以此提出局部中心节点识别算法;将局部中心节点作为种子,计算其对一阶邻居节点的直接吸引力和二阶邻居节点的间接吸引力,将吸引力作为加入社区的判断标准;利用无需设置阈值的重叠节点分析方法对重叠节点进行处理,从而完成社区识别任务.人工基准网络和真实网络的实验结果表明,CDCNCA算法相比其他算法有更好的社区识别效果.
Considering that the detection algorithm of overlapping communities based on central nodes has the shortcoming of instability,we propose an overlapping community detection algorithm based on central nodes and community attraction(CDCNCA).First,we obtain the node centrality coefficient by the node degree and the regional connectivity between a node and its adjacent nodes.On this basis,we propose the local central nodes detection algorithm.Then,the local central nodes are used as seeds to calculate their direct attraction to their first-order neighbor nodes and the indirect attraction to their second-order neighbor nodes.The attractiveness is applied as the criterion for joining the community.Finally,we use the analytical method requiring no threshold setting to process the overlapping nodes to find the overlapping community.The experimental results on artificial benchmark network and real network show that the CDCNCA algorithm has a better community detection effect than other algorithms.
作者
姚继强
邓琨
刘星妍
赵阳
朱莹
凡欣
Yao Jiqiang;Deng Kun;Liu Xingyan;Zhao Yang;Zhu Ying;Fan Xin(College of Mathematics and Computer Science,Zhejiang Normal University,Jinhua,Zhejiang 321004;College of Information Science and Engineering,Jiaxing University,Jiaxing,Zhejiang 314001;Key Laboratory of Medical Electronics and Digital Health of Zhejiang Province,Jiaxing University,Jiaxing,Zhejiang 314001;College of International Studies,Jiaxing University,Jiaxing,Zhejiang 314001;College of Material and Textile Engineering,Jiaxing University,Jiaxing,Zhejiang 314001)
出处
《嘉兴学院学报》
2022年第6期31-41,共11页
Journal of Jiaxing University
基金
教育部人文社会科学研究(高校辅导员)项目(22JDSZ3023)
教育部产学合作协同育人项目(220603372015422,220604029012441)
浙江省教育厅思政专项课题(Y202147298)
浙江省教育科学规划课题项目(2020SCG046)。
关键词
复杂网络
社区识别
中心节点
社区吸引力
complex network
community detection
central node
community attractiveness
