Identifying active subgroups in online communities

ABSTRACT

As online communities proliferate, methods are needed to explore and capture patterns of activity within them. This paper focuses on the problem of identifying active subgroups within online communities. k-plex analysis and hierarchical clustering are used to identify and contrast subgroups, and the methodology is demonstrated in a case study involving the TorCamp Google group community. We assessed the validity of the subgroups obtained in the case study by comparing them with the members' experienced sense of community, and their self-reported acquaintanceships. Results suggest that active subgroups of people not only interact with each other at a higher rate, but also have a greater experienced sense of community. It is concluded that detection of active subgroups in online communities can be implemented widely using automated tools for analyzing the social networks implied by online interactions.

REFERENCES

• 1: C. Bird. Community Structure in OSS Projects. Downloaded from http://wwwcsif.cs.ucdavis.edu/~bird/, July 13, 2007.
• 2: Alvin Chin , Mark Chignell, A social hypertext model for finding community in blogs, Proceedings of the seventeenth conference on Hypertext and hypermedia, August 22-25, 2006, Odense, Denmark [doi>10.1145/1149941.1149945]
• 3: M. Everett. Cohesive subgroups. Analytic Technologies, http://www.analytictech.com/networks/EverettSubgroups.doc
• 4: G. A. Hillery. Definitions of community: Areas of agreement. Rural Sociology, Vol. 20, (1955), pages 779--791.
• 5: J. Hopcroft et al. Tracking evolving communities in large linked networks. PNAS 101: 5249--5253; published online before print as 10.1073/pnas.0307750100.
• 6: S. C. Johnson. Hierarchical Clustering Schemes. Psychometrika, Vol. 2, (1967), pages 241--254.
• 7: D. W. McMillan and D. M. Chavis. Sense of community: A definition and theory. Journal of Community Psychology, Vol. 14, No. 1, (1986), pages 6--23.