An Enhancement for finding User Behavior on Progressive Data Using Weight Constraint

Abstract

Every organization needs to understand their customers' behavior, preferences, and future needs, which often depend on past behavior. Web Usage Mining is an active research topic in which user session clustering is done to understand users' activities. I am proposing an enhancement to the web log mining process based on online navigational pattern prediction. In this paper, I use a neural-based approach, Self-Organizing Mapping (SOM), for clustering sessions as a trend analysis with some parameters. The process depends on the performance of the clustering of the number of requests. Here, I use the SOM algorithm in Frequent Sequential Traversal Pattern Mining, called STPMW. The approach proceeds as follows: first, I use the SOM algorithm to obtain clusters of web logs. Then, we load the web log cluster that is closely related to frequent patterns. After that, I apply the Min-Max Weight of Page in Sequential Traversal Pattern. If the given support lies between the min and max weight range, the item is considered frequent. Otherwise, I check the average weight. Finally, I establish a reliable prediction based on both the quantity of data and the quality of the results.