Quality Enhancement Based on Reinforcement Learning and Feature Weighting for a Critiquing-Based Recommender [chapter]

Maria Salamó, Sergio Escalera, Petia Radeva
2009 Lecture Notes in Computer Science  
Personalizing the product recommendation task is a major focus of research in the area of conversational recommender systems. Conversational case-based recommender systems help users to navigate through product spaces, alternatively making product suggestions and eliciting users feedback. Critiquing is a common form of feedback and incremental critiquing-based recommender system has shown its efficiency to personalize products based primarily on a quality measure. This quality measure
more » ... the recommendation process and it is obtained by the combination of compatibility and similarity scores. In this paper, we describe new compatibility strategies whose basis is on reinforcement learning and a new feature weighting technique which is based on the user's history of critiques. Moreover, we show that our methodology can significantly improve recommendation efficiency in comparison with the state-of-the-art approaches.
doi:10.1007/978-3-642-02998-1_22 fatcat:kxppynacfreh3jxvrtgd4lvx6e