A Road Map to More Effective Web Personalization: Integrating Domain Knowledge with Web Usage Mining

Personalization based on Web usage mining can enhance the effectiveness and scalability of collaborative filtering. However, without semantic knowledge about the underlying domain, such systems cannot recommend different types of complex objects based in their underlying properties and attributes. This paper provides an overview of approaches for incorporating semantic knowledge into Web usage mining and personalization processes. We present two general approaches to integrate semantic knowledge extracted from the content features of pages into the usage-based personalization process. Next, we present a general framework of integrating domain ontologies with Web Usage Mining and Personalization. In each case, we discuss how semantic knowledge is leveraged and represented in the preprocessing and pattern discovery phases, as well as how it is used to enhance usage-based personalization.