An improved optimization of link-based label propagation algorithm
点击次数:
所属单位:计算机科学与技术学院/人工智能学院/软件学院
发表刊物:Lect. Notes Comput. Sci.
摘要:Community Detection has been an important tool for network and overlapping Community exists in real work ubiquitously. In this paper, an improved optimization of link-based label propagation algorithm rather than node called LinkLPAm is proposed to detect overlapping community. We briefly introduce our main work. Firstly, the initialization on edge labels by rough core is presented to speed up the process of detecting overlapping community, which is a big timesaver for the link network that magnified a lot of times compared with original node network. Secondly, an optimization algorithm of label propagation on link is given to update label on edge. Thirdly, in order to restrict the number of communities and the number of nodes in the community, the metric community similarity between communities is defined and greedy mergence algorithm is taken to merge communities according to community similarity. Finally, experimental result shows that our method LinkLPAm is serviceable for find reasonable overlapping community. © 2018, Springer Nature Switzerland AG.
ISSN号:0302-9743
是否译文:否
发表时间:2018-01-01
合写作者:Zhu, Xiaoxiang
通讯作者:夏正友