Enhancing Performance with a Learnable Strategy for Multiple Question Answering Modules

ETRI Journal

关键词: strategy learning;    machine learning;    Question answering;   

Others  :  1185793
DOI  :  10.4218/etrij.09.0108.0388

【 摘 要 】

A question answering (QA) system can be built using multiple QA modules that can individually serve as a QA system in and of themselves. This paper proposes a learnable, strategy-driven QA model that aims at enhancing both efficiency and effectiveness. A strategy is learned using a learning-based classification algorithm that determines the sequence of QA modules to be invoked and decides when to stop invoking additional modules. The learned strategy invokes the most suitable QA module for a given question and attempts to verify the answer by consulting other modules until the level of confidence reaches a threshold. In our experiments, our strategy learning approach obtained improvement over a simple routing approach by 10.5% in effectiveness and 27.2% in efficiency.

   

【 参考文献 】

• [1]J. Chu-Carroll et al., "A Multi-strategy and Multi-source Approach to Question Answering," Proc. 11th Text REtrieval Conference (TREC-11), 2002, pp. 281-288.
• [2]J. Lin and B. Katz, "Question Answering from the Web Using Knowledge Annotation and Knowledge Mining Techniques," Proc. 12th Int. Conf. Information and Knowledge Management, 2003, pp. 116-123.
• [3]B. Katz et al., "Answering Multiple Questions on a Topic from Heterogeneous Resources," Proc. 13th TREC, 2004.
• [4]A. Hickl et al., "Question Answering with LCC’s CHAUCER at TREC 2006," Proc. 15th TREC, 2006.
• [5]D. Moldovan, M. Bowden, and M. Tatu, "A Temporally-Enhanced PowerAnswer in TREC 2006," Proc. 15th TREC, 2006.
• [6]E. Nyberg et al., "The JAVELIN Question Answering System at TREC 2003: A Multi-strategy Approach with Dynamic Planning," Proc. 12th TREC, 2003.
• [7]A. Hickl et al., "Experiments with Interactive Question Answering in Complex Scenarios," Proc. Workshop on the Pragmatics of Question Answering at HLT/NAACL, 2004, pp. 60-69.
• [8]H-J. Oh et al., "An Integration Approach of QA Engines Depending on Answer-Classes," Proc. IEEE IRI, 2006, pp. 178-181.
• [9]C.K. Lee et al., "Fine-Grained Named Entity Recognition Using Conditional Random Fields for Question Answering," Proc. AIRS, vol. 4182, 2006, pp. 581-587.
• [10]C.K. Lee et al., "A Multi-strategic Concept-Spotting Approach for Robust Spoken Korean Understanding," ETRI Journal, vol. 29, no. 2, Apr. 2007, pp. 179-188.
• [11]M.R. Choi, J. Hur, and M-G. Jang, "Constructing Korean Lexical Concept Network for Encyclopedia Question Answering System," Proc. IEEE IECON, vol. 3, Nov. 2004, pp. 3115-3119.
• [12]P.S. Jacobs, G.R. Krupka, and L.F. Rau, "Lexico-semantic Pattern Matching as a Companion to Parsing in Text Understanding," Proc. Workshop on Speech and Natural Language in HLT Conference, 1991, pp. 337-341.
• [13]H.-J. Oh, Compositional Question Answering with Collaborative Strategies, PhD dissertation, Information and Communication Univ. Korea, 2008.
• [14]C.K. Lee and M.G. Jang, "Fast Training of Structured SVM Using Fixed-Threshold Sequential Minimal Optimization," ETRI Journal, vol. 31, no. 2, Apr. 2009, pp. 121-128.
• [15]E.M. Voorhees, "The TREC-8 Question Answering Track Report." Proc. 8th TREC, 1999, pp. 77-82.
• [16]J. Lin. "Is Question Answering Better Than Information Retrieval? A Task-Based Evaluation Framework for Question Series," Proc. HLT/NAACL, 2007, pp. 212-219.
• [17]D. Moldovan et al., "Performance Issuse and Error Analysis in an Open-Domain Question Answering System," ACM Trans. Information Systems (TOIS), vol. 21, no. 2, 2003, pp. 133-154.