Discovering user communities on the Internet using unsupervised machine learning techniques

Abstract

Contribution to special issue on intelligence and interaction in community based systems (Part 1). Argues for the usefulness of constructing communities of users with common behaviour, making use of machine learning techniques. Assumes that the users of any service on the Internet constitute a large community and aims to construct smaller communities of users with common characteristics. Presents the results of 3 case studies for 3 different types of Internet service: a digital library, an information broker and a Web site. Considers the different types of information access involved: query based information retrieval, profile based information filtering and Web site navigation. Evaluates 2 different unsupervised learning methods: conceptual clustering and cluster mining.