International Journal of Advanced Research in Computer Science and Software Engineering (IJARCSSE)
Web log mining can be described as the discovery and analysis of access patterns of users through mining of log files. For analyzing the customer's behavior, the data generated by the users visiting the website must be analyzed. The users' accesses to Web sites are stored in server log files. But the data stored in these log files do not present an accurate picture of the users' accesses to the Web site. So the preprocessing of web log data is a pre-requisite phase before it can be used for mining tasks. The preprocessed web data then is suitable for web mining. This paper presents various steps involved in preprocessing of web log files.