The development and prospect of personalized TV program recommendation systems

Abstract

Aspects of personalized TV program recommendation systems are reviewed. An introduction and relevant work are presented, and philosophies of program description, user profiling and the matching mechanism are illustrated, especially for strategies such as implicit and explicit feedback, and content-based and collaborative filtering techniques. A new system quality evaluation method is then presented. In addition, we propose a next generation conceptual architecture for integrating a real-time program rating mechanism with a personalization system using Web services.