Soft Computing Paradigms for Web Access Pattern Analysis

Abstract

Web servers play a crucial role to convey knowledge and information to the end users. With the popularity of the WWW, discovering the hidden information about the users and usage or access pattern is critical to determine effective marketing strategies and to optimize the server usage or to accommodate future growth. Many of the currently available or conventional server analysis tools could provide only explicit statistical data without much useful knowledge and hidden information. Therefore, mining useful information becomes a challenging task when the Web traffic volume is enormous and keeps on growing. In this paper, we propose Soft Computing Paradigms (SCPs) to discover Web access or usage patterns from the available statistical data obtained from the Web server log files. Self Organising Map (SOM) is used to cluster the data before the data is fed to three popular SCPs including Takagi Sugeno Fuzzy Inference System (TSFIS), Artificial Neural Networks (ANNs) and Linear Genetic Programming (LGP) to develop accurate access pattern forecast models. The analysis was performed using the Web access log data obtained from the Monash University’s central Web server, which receives over 7 million hits in a week. Empirical results clearly demonstrate that the proposed SCPs could predict the hourly and daily Web traffic volume and the developed TSFIS gave the overall best performance compares with other proposed paradigms.