Investigating Internet user Behaviour on Proxy Server Log File Using Web Usage Mining

Abstract

Nowadays, each and all activities have been becoming unimaginable to do without internet. When a web user requests a specific page, an entry is made in a special file known as the server log file. Some organizations use a proxy server for caching services and administrative control. In order to provide the client computer with oblique network services, a proxy server sits in between the client internet and the client's computer. Since proxy server is in between client and web server, every request from clients will pass through the proxy server to corresponding web server. Analysing proxy log file has numerous advantages for access behaviour investigation process. The study's main goal is to investigate internet users' access behaviour by analysing proxy log files. Keeping this in mind, the study was carried out by adhering to the three steps of the online usage mining procedure, including data preparation, pattern finding, and pattern analysis. In pattern discovery phase, data mining methods such as association rule mining and statistical analysis have been used with the intention of discovering patterns. The WEKA data mining tool's Apriori algorithm was employed for association rule mining. The study tried to investigate users’ access behaviour based on two time situations such as class time and exam time. By considering this two time situations, the study answered the research questions. This study will be an input for creating a platform that changes the university internet usage trend in the way to be essential for effective teaching-learning process. In addition, the study will show how analysing web usage mining on proxy server access log data is essential and motivate other researchers.