Relevance ranking for one to three term queries
Information Processing and Management: an International Journal
Performance issues and error analysis in an open-domain question answering system
ACM Transactions on Information Systems (TOIS)
Quantitative evaluation of passage retrieval algorithms for question answering
Proceedings of the 26th annual international ACM SIGIR conference on Research and development in informaion retrieval
Passage retrieval vs. document retrieval for factoid question answering
Proceedings of the 26th annual international ACM SIGIR conference on Research and development in informaion retrieval
COLING '98 Proceedings of the 17th international conference on Computational linguistics - Volume 1
| Hi-index | 0.00 |
In this paper, the effect of using semantic class overlap evidence in enhancing the passage retrieval effectiveness of question answering (QA) systems is tested. The semantic class overlap between questions and passages is measured by evoking FrameNet semantic frames using a shallow term-lookup procedure. We use the semantic class overlap evidence in two ways: i) fusing passage scores obtained from a baseline retrieval system with those obtained from the analysis of semantic class overlap (fusion-based approach), and ii) revising the passage scoring function of the baseline system by incorporating semantic class overlap evidence (revision-based approach). Our experiments with the TREC 2004 and 2006 datasets show that the revision-based approach significantly improves the passage retrieval effectiveness of the baseline system.