Clustering of Preprocessed Web Usage Data Using ART1 Neural Network and Comparative Analysis of ART1, K-Means and SOM Clustering Techniques

Abstract

Web Usage Data is related to web activity, the majority of the techniques that have been used for pattern discovery from Web Usage Data are clustering methods due to their limitations this paper proposes a novel partition based approach for dynamically grouping Web users based on their Web access patterns using ART1 NN clustering algorithm. In e-commerce applications, clustering methods are used for the purpose of generating marketing strategies, product offerings, personalization, web site adaptation and also used for preload web pages which are likely to be accessed in near future.