Hacia el Uso de la Información Sintáctica y Semántica en los Sistema de Búsqueda de Respuesta

Current question answering systems typically combine methods based in information retrieval and information extraction. Such methods leverage the speed of the resulting algorithms and the presence of information redundancy, but few systems use methods based on linguistic information beyond the word level. In this paper we present a study of the impact of the syntactic and semantic information in the selection process of the final candidate sentence. The selection process is based on measures of grammatical relation overlap and flat logical form overlap.