A User Profiles Acquiring Approach Using Pseudo-Relevance Feedback

摘要

User profiles are important in personalized Web information gathering and recommendation systems. The current user profiles acquiring techniques however suffer from some problems and thus demand to improve. In this paper, a survey of the existing user profiles acquiring mechanisms is presented first, and a novel approach is introduced that uses pseudo-relevance feedback to acquire user profiles from the Web. The related evaluation result is promising, where the proposed approach is compared with a manual user profiles acquiring technique.