Prediction mining Generalization-Based clustering method

Mining user patterns of log files can provide significant and useful informative knowledge. This paper present an approach for mining similarity of interest among web users from their past access behaviors. Unlike traditional clustering methods that focus on grouping objects with similar values on a set of dimensions, clustering by pattern similarity finds objects that exhibit a coherent pattern of rise and fall in subspaces. The similarity measures are based on the user sessions extracted from the user’s access logs. The method will be evaluated its use for usage characterization based on a hierarchical clustering method.