Sequence Association Rule with Pattern Constraints in Web Usage Mining: data stream mining

Sequence Association rule prediction is one of the most important activity in web usage mining. In this dissertation work the data in web log files are preprocessed. The irrelevant data (images and intermediate) in the log files are removed for further processing. The processed data is then analyzed for user and session identification. Finally the data in the form of session sequences are analyzed to find the frequent pattern. The pattern is then considered for the rule prediction.