An Improved Session Identification Approach in Web Log Mining for Web Personalization

This Web based applications are increasing at an enormous speed and as a result its users are also increasing at an exponential speed. The innovative and evolutionary changes in technology have made it possible to capture the user's fundamental nature and interactions with web applications through web server log file as web usage. In order to design attractive web sites, designers must understand their user's needs. Therefore analyzing navigational behavior of users is an important part of web page design. Web Usage Mining (WUM) is the application of data mining techniques to web usage data in order to discover the patterns that can be used to analyze the user's navigational behavior. Since web contains large amount of "irrelevant information" in the web log, the original log file cannot be directly used in the WUM process. Therefore, the preprocessing of web log file becomes very important in order to improve the accuracy in Web log mining. The basic procedure of data preprocessing is introduced firstly in this paper with the traditional session identification algorithm is fully analyzed, on the basis of which a session identification algorithm on page threshold and dynamic timeout is presented. Finally, the initial timeout is computed for each page according to sessions formed, combining with the importance degree of improved dynamic threshold algorithm which discards the uninterested attributes from log file. Comparing experiment shows that the algorithm Proposed can obtain a better performance on session identification and user interests which is the key for web personalization.