|
Title:
|
COMPARISON OF COMMUNITY IDENTIFICATION TECHNIQUES FOR TWO-MODE AFFILIATION NETWORKS USING WIKIPEDIA DATA |
|
Author(s):
|
Fawad Nazir , Hideaki Takeda , Aruna Seneviratne |
|
ISBN:
|
978-972-8924-78-2 |
|
Editors:
|
Piet Kommers and Pedro Isaías |
|
Year:
|
2009 |
|
Edition:
|
1 |
|
Keywords:
|
Social Networks, Community Identification, Cohesive subgroups, Edge Betweenness, Hierarchical Clustering,
Dendrograms. |
|
Type:
|
Full Paper |
|
First Page:
|
439 |
|
Last Page:
|
445 |
|
Language:
|
English |
|
Cover:
|
|
|
Full Contents:
|
click to dowload 
|
|
Paper Abstract:
|
One of the most important questions in social networks is the identification of cohesive subgroups (a.k.a. community
identification). These cohesive subgroups are loosely defined as collection of individuals who interact frequently. Once
the communities are identified they often reveal interesting properties of the social network members, such as common
hobbies, interests, social bindings, occupations etc. Several types of algorithms exist for analysis and identification of
cohesive subgroups in one-mode networks that focus on pair-wise ties. However, less attention has been given to
identification of cohesive subgroups in two-mode affiliation networks. Two mode affiliation networks focus on ties
existing among actors through joint affiliations. Therefore, in this paper we evaluate two cohesive subgroups
identification methods i.e. edge betweenness and hierarchical clustering, for two-mode affiliation network using the
Wikipedia data. We conclude from our results that edge betweenness technique, when applied to two-mode affiliation
network, is a better techniques in terms of the modularity value that means it can generate more strong social
communities in terms of social ties. On the other hand this technique is less time efficient as compared to hierarchical
clustering. |
|
|
|
|
|
|
