Paper Title
Refined Clusters for the Improvement of Quality in Terms of Web Recommendation

Abstract
The information overload problem demands the need for refinement of Web Usage Clusters for web based recommendations and classified learning. Web Usage Clusters have a limitation of poor quality. To overcome these limitations, cluster refinement framework has been proposed in this paper to generate clusters with improved quality along with improved accuracy of recommendations to the target users and classified learning for web page predictions. Keywords - Web Usage Clusters, Knockout Refinement Algorithm (KRA)