A graph-based web usage mining method considering client side data

Abstract

To improve Web site, we need to evaluate current usage of it. Web usage mining and statistical analysis are two ways to evaluate usage of Web site. The combination of Web usage mining and statistical analysis gives more accurate information about Web usage. Through Web usage mining methods, graph mining covers complex Web browsing behaviors such as parallel browsing. Through statistical analysis methods, analyzing page browsing time gives valuable information about Web site and its users. This paper presents a Web usage mining method which combines Web usage mining and statistical analysis considering client side data. In other words, it combines graph based Web usage mining and browsing time analysis with taking client side data into account. It helps us to reconstruct user session exactly as it has been and based on these data, we find Web usage patterns with more accuracy.