ABSTRACT

Web-based communities, such as those created around weblogs, form increasingly complex networks, and new tools are needed to map and understand them. Creating a community map allows the visualization of community standing and relationship, and it can be used to discover which members of the community have similar interests. Since hyperlinks are an indication of a member interests, the set of all hyperlinks by a member can be used to represent it; this set can then be used to perform clustering procedures on the group of all community members. In this paper, such procedure is carried out using Kohonen's self-organizing map (SOM), a neural-net like method used to create maps. SOM can be used to divide the set of webs under study in communities/clusters, and, at the same time, visualize them, so that a map for community navigation can be created out of the initial map. This procedure has been applied to the Blogalia weblog community (hosted at http://www.blogalia.com/), a thriving community of around 200 members, created in January 2002. In this paper we show how SOM discovers interesting community features, as well as its possible shortcomings when mapping communities