Date Added: Jun 2010
Clustering is useful technique in the field of textual data mining. Cluster analysis divides objects into meaningful groups based on similarity between objects. Copious material is available from the World Wide Web (WWW) in response to any user-provided query. It becomes tedious for the user to manually extract real required information from this material. This paper proposes a scheme to effectively address this problem with the help of cluster analysis. In particular, the ROCK algorithm is studied with some modifications. ROCK generates better clusters than other clustering algorithms for data with categorical attributes.