融合节点属性的局部多重社区发现算法  

Multiple local community detection with integrated node attributes

作　　者：陈李舟 冯俊又 徐煊翔 刘先博 杜彦辉[1] Chen Lizhou;Feng Junyou;Xu Xuanxiang;Liu Xianbo;Du Yanhui(College of Information&Cyber Security,People’s Public Security University of China,Beijing 100038,China)

机构地区：[1]中国人民公安大学信息网络安全学院,北京100038

出　　处：《计算机应用研究》2025年第1期42-47,共6页Application Research of Computers

基　　金：中国人民公安大学网络空间安全执法技术双一流创新研究专项资助项目。

摘　　要：局部多重社区发现是社交网络分析中的关键技术,旨在揭示网络中用户的多重归属和复杂联系。针对现有局部多重社区发现算法大多基于网络拓扑结构,忽视节点属性信息的问题,提出了融合节点属性的局部多重社区发现算法(MLCDINA)。该算法将属性网络的结构和属性信息相结合为节点对之间的边权重,并通过随机游走评估节点间结构和属性的融合重要性(IISA)。此外,该算法引入了考虑边权重的局部聚类系数和亲密度随机游走(IRW),以增强对子图稠密性和IISA的评估。实验结果表明,MLCDINA在真实属性网络上的Jaccard F 1-score较现有算法有显著提升,验证了其在局部多重社区发现任务中的有效性。Multiple local community detection is a key technology in social network analysis,aiming to reveal the multiple affiliations and complex connections of users within networks.Addressing the issue that most existing multiple local community detection algorithms are based on network topology and neglect node attribute information,this paper introduced an algorithm named multiple local community detection with integrated node attributes(MLCDINA).This algorithm combined the structure and attribute information of the attributed network to determine the edge weights between node pairs and evaluated the importance of the integration of structure and attributes(IISA)through random walks.In addition,the algorithm introduced a local clustering coefficient that considered edge weights and an intimacy random walk(IRW)to enhance the evaluation of subgraph density and IISA.Experimental results indicate that MLCDINA significantly improves the Jaccard F 1-score over existing algorithms on real attributed networks,verifying its effectiveness in multiple local community detection tasks.

关 键 词：局部社区发现 属性网络 随机游走 

分 类 号：O157.5[理学—数学]