CONICET | Buscador de Institutos y Recursos Humanos

Several techniques for improving websearch have been developed over the last years.Most existing approaches are stilllimited, mainly due to the absence~of qualitative criteria forranking results and insensitivity to user preferences for guidingthe search. At the same time, defeasible argumentation evolved asa successful approach in AI to model commonsense qualitativereasoning with applications in many areas, such as agent theory,knowledge engineering and legal reasoning. This paper presents\textsc{ArgueNet}, a recommender system that classifies searchresults according to preference criteria declaratively specifiedby the user. The proposed approach integrates a traditional websearch engine with a defeasible argumentation framework.

enviar mensaje