Dynamic Modeling by Usage Data for Personalization Systems

With the extensive growth of data available on the Internet, personalization of this huge information becomes essential. Although, there are various techniques of personalization, in this paper we concentrate on using data mining algorithms to personalize web sitespsila usage data. This paper proposes an off-line model based web usage mining that is generated by clustering algorithm.Then, we will use userspsila transactions periodically to change the off-line model to a dynamic-model.This proposed approach will solve the problem of the decrease of accuracy in the off-line models over time resulted of new users joining or changes of behaviour for existing users in model-based approaches. Finally, we discuss the on-line model for user behaviour prediction in the web personalization system.