Pre-proccesing log data for User Behavior Analysis

User behavior modeling is one of web usage mining applications helping to understand the behavior of a user while interacting with the Web. Since the activity information of users are recorded in web log files, log data is the starting point for any web mining process. Thus, the quality of results strongly depends on the quality of input data. In this paper, the result of our scenario-based experiment in preprocessing log data for analyzing user behavior is presented. The proposed method is practically applied on Ferdowsi University of Mashhad (FUM) log data and the results are discussed.