Personalized Context-aware Access to Digital Cultural Heritage

This paper presents a semantic recommender method and a system for a personalized access to digital cultural heritage through context-aware user pro- filing. Given annotation knowledge-bases, explicit backgr ound knowledge in the form of ontologies, a user model capturing the user's behavi or and context, the system produces recommendations. Ontology-based user profiling can be used to reduce cold-start, sparsity and over-specialization pr oblems. In addition, we present a recommendation retrieval method that is based on the vector space model and uses indices that enable fast and scalable implementation of the sys- tem.