Cluster Optimization for Improved Web Usage Mining

Abstract

Now days, World Wide Web (WWW) has become rich and most powerful source of information. Conversely, it has become tricky and critical task to retrieve actual information due to its continuous expansion in dimensions. Web Usage Mining is a step-wise technique of extracting useful access patterns of the user from web. Web personalization makes use of web usage mining techniques, for knowledge acquisition process done by analyzing the user navigational patterns. The web page personalization involves clustering of different web pages having similar navigation patterns for an individual. Since cluster size expands due to the frequent access, optimization or shrinking the size of clusters becomes a chief consideration. This paper proposes a tactic of cluster optimization based on concept of swarm intelligence techniques. Later on based on the recognition of user access patterns, clustering is implemented using neural fuzzy approach i.e. NEF Class algorithm and cluster optimization is implemented using Ant Nest Mate Approach.