一种新的复合核函数及在问句检索中的应用

王君; 李舟军; 胡侠; 胡必云

doi:10.3724/SP.J.1146.2010.00268

一种新的复合核函数及在问句检索中的应用

doi: 10.3724/SP.J.1146.2010.00268

王君^{*① 李舟军,},
李舟军,
胡侠,
胡必云

基金项目:

国家973规划项目( 2007CB310803)资助课题

计量
- 文章访问数: 3571
- HTML全文浏览量: 93
- PDF下载量: 970
- 被引次数: 0
出版历程
- 收稿日期: 2010-03-23
- 修回日期: 2010-07-05
- 刊出日期: 2011-01-19

A Novel Composite Kernel and Application to Question Retrieval

Wang Jun^{*① 李舟军
,},
Li Zhou-Jun,
Hu Xia,
Hu Bi-Yun

摘要

摘要: 问句检索在问答系统中有着重要的作用，其核心问题在于研究查询问句与候选问句之间的相似性计算问题，实现问句之间的高精度匹配。该文采用树核函数的方法计算问句之间的结构相似性，并针对原有算法的不足，做了相应的改进。为降低句法解析器性能对树核函数的影响，该文在改进的树核函数基础上，将其与字符串核结合，提出了一种能同时融合问句的句法信息，词性信息和词序信息的复合核函数，用以计算问句之间的综合语义相似性。在社区问答系统Yahoo!Answer的数据上进行测试，相对传统的基于词频的特征向量法，问句检索平均准确率提高了24.02%。
- 信息检索 /
- 问答系统 /
- 问句检索 /
- 复合核函数
Abstract: Question retrieval plays important role in question and answering systems. The main problem is how to measure the similarity between candidate questions and query question. This paper presents a tree kernel based method, named weighted tree kernel, to calculate the similarity of sentences structures and proposes improvements to the original tree kernel algorithm. In order to reduce the effect on tree kernel bringing by syntactic parsing, a composite kernel is proposed based on the weighted tree kernel and two other string kernels, which can capture syntax, part-of-speech and lexical level information of a sentence, to calculate the semantic similarity between question sentences. Experimental results on Yahoo!Answers dataset show that the proposed method outperforms traditional vector space model based methods by 24.02% in question retrieval accuacry.
- Information retrieval /
- Question answering system /
- Question retrieval /
- Composite kernel

HTML全文

参考文献(1)

Burke R D, Hammond K J, and Kulyukin V A, et al.. Question answering from frequently asked question files: experiments with the faq finder system[J]. AI Magazine, 1997, 18(2): 57-66.[2]Jijkoun V and De Rijke M. Retrieving answers from frequently asked questions pages on the web [C]. In CIKM05: Proceedings of the 14th ACM international conference on Information and knowledge management, Bremen, Germany, 2005: 84-90.[3]Cao Xin.[J].Cong Gao, and Cui Bin, et al.. The use of categorization information in language models for question retrieval [C]. In CIKM09: Proceeding of the 18th ACM conference on Information and knowledge management, Hong Kong, China.2009,:-[4]Duan Hui-zhong, Cao Yun-bo, and Lin Chin-yew, et al.. Searching questions by identifying questions topic and question focus [C]. In ACL-08: HLT: Proceeding of the 46th annual meeting of the association for computational linguistics: Human Language Technologies. Columbus, OH, USA, 2008: 156-164.Xue Xiao-bing, Jeon J, and Croft W B. Retrieval models for question and answer archives [C]. In SIGIR '08: Proceedings of the 31st annual international ACM SIGIR conference on Research and development in information retrieval, New York, NY, USA, 2008: 475-482.[5]Wang Kai.[J].Ming Zhao-yan, and Chua Tat-seng. A syntactic tree matching approach to finding similar questions in community-based QA services [C]. In SIGIR09: Proceedings of the 32nd international ACM SIGIR conference on Research and development in information retrieval, Boston, MA, USA.2009,:-[6]Collins M and Duffy N. Convolution Kernels for Natural Language [M]. Advances in Neural Information Processing Systems 14, MIT press, 2001: 625-632.[7]Zhao Shu-bin and Grishman R. Extracting relations with integrated information using kernel methods [C]. In ACL05: Proceedings of the 43rd Annual Meeting on Association for Computational Linguistics, Ann Arbor, Michigan, USA, 2005: 419-426.[8]Lodhih H, Saunders G, and Shawe-Taylor J, et al.. Text classification using string kernels [J].Journal of Machine Learning Research.2002, 2(Feb):419-444[9]Choon Hui-teo and Vishwanathan S V N. Fast and space efficient string kernels using suffix arrays [C]. In ICML06: Proceedings of the 23rd international conference on Machine learning, Pittsburgh, Pennsylvania, USA, 2006: 929-936.[10]Cancedda N, Gaussier E, and Goutte C, et al.. Word sequence kernels [J].The Journal of Machine Learning Research.2003, 3(Feb):1059-1082[11]Joachims T, De Thorsten J, and Cristianini N, et al.. Composite kernels for hypertext categorization. In ICML: International Conference on Machine Learning, Williams College, USA, 2001: 250-257.

施引文献

资源附件(0)

访问统计