Deakin University

File(s) under permanent embargo

Discovering and tracking active online social groups

conference contribution
posted on 2017-01-01, 00:00 authored by Md Musfique Anwar, Chengfei Liu, Jianxin LiJianxin Li, Tarique Anwar
Most existing works on detection of social groups or communities in online social networks consider only the common topical interest of users as the basis for grouping. The temporal evolution of user activities and interests have not been thoroughly studied to identify their effects on the formation of groups. In this paper, we investigate the problem of discovering and tracking time-sensitive activity driven user groups in dynamic social networks. The users in these groups have the tendency to be temporally similar in terms of their activities on the topics of interest. To this end, we develop two baseline solutions to discover effective social groups. The first solution uses the network structure, whereas the second one uses the topics of common interest. We further propose an index-based method to incrementally track the evolution of groups with a lower computational cost. Our main idea is based on the observation that the degree of user activeness often degrades or upgrades widely over a period of time. The temporal tendency of user activities is modelled as the freshness of recent activities by tracking the social streams with a fading time window. We conduct extensive experiments on two real data sets to demonstrate the effectiveness and performance of the proposed methods. We also report some interesting observations on the temporal evolution of the discovered social groups.



Web Information Systems Engineering. International Conference (2017 : Puschino, Russia)




Lecture Notes in Computer Science


59 - 74




Puschino, Russia

Place of publication

Cham, Switzerland

Start date


End date






Publication classification

E1.1 Full written paper - refereed

Copyright notice

2017, Springer International Publishing AG

Title of proceedings

WISE 2017: International Conference on Web Information Systems Engineering