---

Recommender Systems are the common solution to the problem of content overload. This book presents a novel content-based recommender enhanced with semantic knowledge, which can overcome the main limitations of Collaborative Filtering approaches related to the lack of user data: the cold-start and the data sparsity. The main novelties of the proposed recommender are: (1) the user-profile learning algorithm, which combines user´s feedback from different channels and employs specific domain inferences to construct accurate user profiles; (2) the prediction method, which exploits the semantic structure of the ontologies to generate accurate predictions. The system´s design proposed is flexible enough to be potentially applied to Web applications of any domain that can be properly described using ontologies based on Semantic Web technologies. As case study, the proposed recommender has been integrated into an existing tourism application that provides information about tourist attractions. This book should be especially useful to professionals in Computer Science field who may be interested in developing a recommender system able to exploit explicit domain knowledge about items.

---

He is a researcher in the Computer Science Department at the Universitat Politècnica de Catalunya (UPC), Spain. His research interests focus on the study of recommendation approaches that can exploit the available domain semantics in order to improve the prediction performance.

---