作者: Lahsen Abouenour , Paolo Rosso , Karim Bouzouba
DOI:
关键词:
摘要: This paper describes an approach for improving the re- ranking of passages Arabic Question/Answering (Q/A). implements a process performing semantic Query Expansion (QE) based on WordNet (AWN) ontology with structure-based Passage Retrieval (PR) Distance Density n-gram model. Experiments set translated CLEF and TREC questions have shown that accuracy, Mean Reciprocal Rank number answered been significantly improved using our approach. These experiments conducted in open domain by querying web as document collection. An analysis obtained performances is discussed this paper.