用于社区发现的LPA_LRDC标签传播算法  

LPA_LRDC Tag Propagation Community Discovery Algorithm Based on Optimized LeaderRank

作　　者：徐成林[1] 陈志刚[1] 黄瑞[1] 龙增艳 李博[1] 周清清[1] 邓伊琴 

机构地区：[1]中南大学软件学院,长沙410075

出　　处：《小型微型计算机系统》2017年第8期1746-1750,共5页Journal of Chinese Computer Systems

基　　金：国家自然科学基金项目(61379057)资助;中南大学中央高校基本科研业务费专项资金项目(2016zzts368)资助

摘　　要：由于传统的LPA算法,在节点标签更新的顺序以及标签传播过程中存在较大的随机性,给社区发现的准确性和稳定性造成了很大的影响.本文提出LRDC(Leader Rank algorithm considered degree and clustering coefficient)算法并用其来衡量节点的重要性,然后按照节点的重要性大小排序作为LPA算法中初始化节点标签的依据,并在标签传播过程中综合考虑节点重要性以及邻居标签的数量提出LPA_LRDC(Label Propagation Algorithm based on LRDC)标签传播社区发现算法.通过在人工和真实的网络数据集上的实验结果表明,本文提出的标签传播社区发现算法能够显著的提高社区发现的准确性和稳定性.Because of the large random in the order of node label updating and the label propagation process in the traditional LPA algorithm, the accuracy and stability of the community discovery is influenced largely. In this paper, the LRDC （LeaderRank algorithm considered degree and clustering coefficient ） algorithm is proposed and used to measure the importance of nodes, and the nodes sorted according to the importance of node are the basis for the initialization of node labels in the LPA algorithm, and the importance of nodes and the number of neighbor tags are considered in the process of tag communication, and then the LPA_LRDC （ Label Propagation Algorithm based on LRDC） tag communication community discovery algorithm is proposed. Experimental results on both artificial and real network data sets show that the proposed algorithm can significantly improve the accuracy and stability of community detection.

关 键 词：标签传播 LEADER RANK 节点重要性 社区划分 社区发现 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]