A multi-user perspective for personalized email communities
作者：
Highlights：
• We introduced the notion of multi-user personalized email communities.
• We provide the evolution of communities in terms of network properties.
• It helps email strainer, group prediction, and fraudulent account detection.
• It has the potential to be integrated with online email visualization tools.
• To approximate the entire network communities with restricted information.
摘要
•We introduced the notion of multi-user personalized email communities.•We provide the evolution of communities in terms of network properties.•It helps email strainer, group prediction, and fraudulent account detection.•It has the potential to be integrated with online email visualization tools.•To approximate the entire network communities with restricted information.
论文关键词：Graph clustering,Community detection,Collaborative similarity,K-Medoid clustering,Entropy,Density,F-measure,Personalized,Multi-user,Email network,Social media,PI-Net
论文评审过程：Received 5 December 2014, Revised 28 January 2016, Accepted 29 January 2016, Available online 8 February 2016, Version of Record 22 February 2016.
论文官网地址：https://doi.org/10.1016/j.eswa.2016.01.046
