In offline,WEB mining tasks can be executed in the lo gs of WEB service resulting in a user transaction file,and the frequent user t ransaction patterns are extracted by filtering w.

First,analyzing user transaction to find comparability matrix,then by the comparability of user transaction comparability and user\'s scanning path this method makes an average.

Research of a user clustering model based on genetic algorithm;

Application of user clustering based on interest in intrusion detection;

The paper presents an ant colony clustering algorithm based on directional similarity according to Web data characteristic of high-dimension and applys to Web user clustering.

The paper performs users clustering combining association rules with FCM (Fuzzy C-Means), firstly it uses the transaction of users accessing to construct the association matrix on pages, and calculates the similarity degree of the user access paths based on that, then it uses the FCM algorithm to perform users clustering.

The system uses transaction clustering,usage clustering,and association rule discovery to extract usage knowledge from server log file, and combine with the current status of an ongoing Web activity to perform real-time personalization.