Web Users Clustering Based on Interest Degree

Web users' clustering based on interest degree is proposed in this paper,which is based on time aware web user clustering.Users' interest degree is computed in our approach through analyzing users' access patterns in log files,which is adjusted and updated by gradual forgetting algorithm.Through the updated users' interests we build users' clustering.Experiment shows that our approach can analyze users' access patterns better,compute users' interest more precisely,and has better clustering quality.