Web Access Data Visualization with Data Analysis Methods

Abstract

In the course of a travel around the interne, users visit various host servers located all over the world and, inadvertently, place digital traces of their surfing habits. These hosts gather and register information about all users and their client computers for various reasons, such as security violation detection, frequency of page access and browser types. This paper shows how Correspondence Analysis and Hierarchical Classification, two popular data analysis methods can be utilized with web log statistics for the examination of internet user activities and access patterns. We plotted a web access data set from a university department on the first factorial plane and employed a clustering dendrogram to clarify group patterns. We believe that these two data analysis methods, when applied on web access data can aid both web designers and design theorists to attain a better picture on client habits and assess the usability and effectiveness of their design choices. Findings of such a methodology can also be applied in marking, risk management and customer relationship management.