局部扩展的标签传播蚁群优化重叠社区发现  

Local extended label propagation ant colony optimization overlapping community detection

作　　者：李有红[1] 战荫伟[2] 王学军[1] 刘若冰[3] Li Youhong;Zhan Yinwei;Wang Xuejun;Liu Ruobing(Huali College,Guangdong University of Technology,Guangzhou 511325,China;School of Computer Science ＆ Technology,Guangdong University of Technology,Guangzhou 510006,China;College of Arts ＆ Sciences,Yunnan Normal University,Kunming 650224,China)

机构地区：[1]广东工业大学华立学院,广州511325 [2]广东工业大学计算机学院,广州510006 [3]云南师范大学文理学院,昆明650224

出　　处：《计算机应用研究》2018年第7期2071-2074,2078,共5页Application Research of Computers

基　　金：广东高校省级重点平台和重大科研项目(青年创新人才类)(2015KQNCX221);广东省科技计划项目(2013B090500038)

摘　　要：为了提升社区发现算法速度和高精度命题,提出一种局部扩展标签传播算法与蚁群优化结合的重叠社区发现框架(ELPA-ACO)。利用局部扩展的标签传播社区发现算法快速获得蚁群初始信息素和位置;结合网络中节点的拓扑结构、内部标签属性、历史信息和节点的传播相互影响力等因子改进蚂蚁转移概率,进一步提高划分精度。在未知网络和真实网络上验证,ELPA-ACO算法可适应多种网络划分,无论是速度还是精度都有显著提高。In order to improve the speed and accuracy of the algorithm,this paper proposed an overlapping community detection framework based on ELPA-ACO. Firstly,it utilized the extended local label to propagate community detection algorithm to get the initial pheromone of ant colony and position. Then,it improved ant transition probability,combined with the spread of nodes in the network topology,internal attribute,and historical information and node mutual influence factors,to further increased the classification accuracy. In the unknown network and real network,ELPA-ACO algorithm can adapt to a variety of network community detection,both speed and accuracy are significantly improved.

关 键 词：重叠社区发现 蚁群优化算法 网络局部传播特性 局部扩展标签传播算法 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]