RESEARCH ON PASSAGE RETRIEVAL USING DOMAIN KNOWLEDGE IN CHINESE QUESTION ANSWERING SYSTEM

摘要：

In order to improve the accuracy of passage retrieval in restrict domain question answering system, The paper proposes a retrieval method using domain knowledge in Chinese question answering system. This method extracts question keywords by analyzing syntactic structure, expands synonymous keywords with the help of HowNet and Field HowNet to do question query, and then extracts text with improved NUTCH retrieval system, finally improves the retrieval method MuliTex according to domain knowledge, arranges the order and calculates weight of paragraphs. On this base, the method extracts related paragraphs as answers, and then has carried on the question and answer experimental verification in the traveling domain. The result shows this method has a good effect. The value of retrieval MRR is up to 0.45.

关键词： Question answering system Passage retrieval Domain knowledge

作者: LU HAN ZHENG-TAO YU YAN-X1A QIU XIANG-YAN MENG JIAN-YI GUO SHENG-TAO SI

作者单位: The School of Information Engineering and Automation, Kunming University of Science and Technology, The School of Information Engineering and Automation, Kunming University of Science and Technology, The School of Information Engineering and Automation, Kunming University of Science and Technology,

会议类型: 国际会议

会议名称: 2008 International Conference on Machine Learning and Cybernetics(2008机器学习与控制论国际会议)

会议地点: 昆明

会议语种:英文

页码: 2603-2606

在线出版日期: 2008-07-12（万方平台首次上网日期，不代表论文的发表时间）