Web usage mining: An implementation view

Abstract

This paper describes the implementation of Web usage mining for DSpace server of NIT Rourkela. The DSpace log files have been preprocessed to convert the data stored in them into a structured format. Thereafter, the general procedures for bot-removal and session-identification from a Web log file have been applied with certain modifications pertaining to the DSpace log files. Furthermore, analysis of these log files using a subjective interpretation of recently proposed algorithm EIN-WUM has also been conducted. © 2011 Springer-Verlag Berlin Heidelberg.