TH-PASS: PERSONALIZED ACTIVE SERVICE SYSTEM IN DIGITAL LIBRARY

摘要

In this paper, as a solution to the problem of Information overload in digital library, we present our research on building TH-PASS-- a personalized active service system. TH-PASS is able to provide the users personalized filtering and active notification service based on user modeling and user pattern learning/discovering. The system's performance is improved, by making use of relevance measure of multiple features including query keywords, term, citation vectors, and user action parameters. The framework of TH-PASS is proposed, and the user model and profile learning are discussed.