Overlapping community detection by constrained personalized PageRank
作者：
Highlights：
• We propose constrained personalized PageRank for local community detection.
• We propose a method to compute an approximate constrained personalized PageRank.
• We give a local optimal version of a sweep that produces connected communities.
• We propose a sequence-based seeding method for local community detection.
• Our community detection method achieves the best accuracy on large real networks.
摘要
•We propose constrained personalized PageRank for local community detection.•We propose a method to compute an approximate constrained personalized PageRank.•We give a local optimal version of a sweep that produces connected communities.•We propose a sequence-based seeding method for local community detection.•Our community detection method achieves the best accuracy on large real networks.
论文关键词：Local community detection,Constrained personalized PageRank,Random walk,Redundant diffusion
论文评审过程：Received 24 February 2020, Revised 19 July 2020, Accepted 3 February 2021, Available online 10 February 2021, Version of Record 23 February 2021.
论文官网地址：https://doi.org/10.1016/j.eswa.2021.114682
