Perception based User Profiles for Web Personalization

Abstract

Personalized web services reduce the burden of information overload by collecting facts that match the needs of the user. An important aspect of personalized web services is the creation of user profiles that contain user information and settings. This article introduces a unique method called Perception-Based User Profiles (PUP) based on perception and browsing order, develops and updates user profiles. User profiles include perceptions and relationships, which can help guarantee that user interests are represented semantically. Second, when calculating the perception and duration of the relationship, for each site in a session, the user's browsing order is considered. Third, cognitive psychometric memory model is used to update the user profile's perceptions and relationships at the end of each session, ensuring the user profile's dynamics. The results of the tests suggest that this strategy works well for building and updating user profiles.