Identifying subcommunities using cohesive subgroups in social hypertext

  • Authors:
  • Alvin Chin;Mark Chignell

  • Affiliations:
  • University of Toronto, Toronto, ON, Canada;University of Toronto, Toronto, ON, Canada

  • Venue:
  • Proceedings of the eighteenth conference on Hypertext and hypermedia
  • Year:
  • 2007

Quantified Score

Hi-index 0.00

Visualization

Abstract

Web pages can be modeled as nodes in a social network, and hyperlinks between pages form links (relationships) between the nodes. Links may take the form of comments, for example on blogs, creating explicit connections between authors and readers. In this paper, we describe a novel methodology and framework for identifying subcommunities as cohesive subgroups of n-cliques and k-plexes within social hypertext. We apply our methodology to a group of computer technologists in Toronto called TorCamp who communicate using a Google group. K-plex analysis is then used to identify a group of people that forms a subcommunity within the larger community. The results are then validated against the experienced sense of community of people inside and outside the subcommunity. Statistically significant differences in experienced sense of community are found, with people within the subcommunity showing higher levels of perceived influence and emotional connection.