Exploiting Ontological Structure for Complex Preference Assembly

When a user is looking for a product recommendation they usually lack expert knowledge regarding the items they are looking for. Ontologies on the other hand are crafted by experts and therefore provide a rich source of information for enhancing preferences. In this paper we significantly extend previous work on exploiting ontological information by allowing the user to specify preferences in a more expressive manner. Rather than allowing for only one preferred target concept, we allow a `chain' of user preferences. Furthermore, we treat information from the underlying ontology of the domain as a secondary preference structure. We then show how to assemble these two preference structures (user and ontology) into a preference over items. The ability to model preferences and exploit preferential information to assist users in searching for items has become an important issue in artificial intelligence. Accurately eliciting preferences from the user in the form of a query can result ...

• Type: Conference
• Year: 2008
• Where: AUSAI
• Authors: Gil Chamiel, Maurice Pagnucco
• Keywords: Artificial Intelligence, AUSAI 2008, Expert Knowledge, Preference Structures, User