Persistent Community Detection in Dynamic Social Networks

• Main Authors: LIU, Siyuan, WANG, Shuhui, KRISHNAN, Ramayya
• Format: text
• Language: English
• Published: Institutional Knowledge at Singapore Management University 2014
• Subjects: Community detection; persistent behavior; social networks; Computer Sciences; Theory and Algorithms
• Online Access: https://ink.library.smu.edu.sg/sis_research/3479; https://ink.library.smu.edu.sg/context/sis_research/article/4480/viewcontent/C101___Persistent_Community_Detection_in_Dynamic_Social_Networks__PAKDD2014_.pdf
• Institution: Singapore Management University

While community detection is an active area of research in social network analysis, little effort has been devoted to community detection using time-evolving social network data. We propose an algorithm, Persistent Community Detection (PCD), to identify those communities that exhibit persistent behavior over time, for usage in such settings. Our motivation is to distinguish between steady-state network activity, and impermanent behavior such as cascades caused by a noteworthy event. The results of extensive empirical experiments on real-life big social networks data show that our algorithm performs much better than a set of baseline methods, including two alternative models and the state-of-the-art.