Web usage mining using data mining methods and techniques

Abstract

Web usage mining aims at utilising data mining as well as other computational methods to analyse and process web server logs that portray the web usage activities of online users In this context the relevant data from web server logs are extracted and transformed to meet the requirements of the computational algorithms The learning models built from computational algorithms help in discovery of interesting patterns that aid in enhancing the web environment Web cache optimization and web personalization are the two most significant concerns in the field of web research Significant web patterns viz subsequently accessed interesting page categories and user groups targeting improved web cache optimization and web personalization are focussed in this research Prediction of subsequently accessed interesting page categories help in improved design of web caches Formation of user groups provides insights into the interests of users thereby helping in enhanced web customization Benchmark dataset viz MSNBC dataset available at UCI repository has been used for validation of methods towards elicitation of web patterns The dataset contains web page navigation sequences of various users newline