|Table of Contents|

[1] Yang Sichun, Gao Chao, Yao Jiamin, et al. Feature combination via importance-inhibition analysis [J]. Journal of Southeast University (English Edition), 2013, 29 (1): 22-26. [doi:10.3969/j.issn.1003-7985.2013.01.005]
Copy

Feature combination via importance-inhibition analysis()
基于重要性和抑制性分析的问句特征组合
Share:

Journal of Southeast University (English Edition)[ISSN:1003-7985/CN:32-1325/N]

Volumn:
29
Issue:
2013 1
Page:
22-26
Research Field:
Computer Science and Engineering
Publishing date:
2013-03-20

Info

Title:
Feature combination via importance-inhibition analysis
基于重要性和抑制性分析的问句特征组合
Author(s):
Yang Sichun1 2 Gao Chao3 Yao Jiamin2 Dai Xinyu1 Chen Jiajun1
1State Key Laboratory for Novel Software Technology, Nanjing University, Nanjing 210093, China
2School of Computer Science, Anhui University of Technology, Maanshan 243032, China
3School of Computer Science and Information Engineering, Chuzhou University, Chuzhou 239000, China
杨思春1 2 高超3 姚佳岷2 戴新宇1 陈家骏1
1南京大学计算机软件新技术国家重点实验室, 南京 210093; 2安徽工业大学计算机学院, 马鞍山 243032; 3滁州学院计算机与信息工程学院, 滁州 239000
Keywords:
question answering system question classification feature combination importance-inhibition analysis
问答系统 问题分类 特征组合 重要性和抑制性分析
PACS:
TP391
DOI:
10.3969/j.issn.1003-7985.2013.01.005
Abstract:
A new method for combining features via importance-inhibition analysis(IIA)is described to obtain more effective feature combination in learning question classification. Features are combined based on the inhibition among features as well as the importance of individual features. Experimental results on the Chinese questions set show that, the IIA method shows a gradual increase in average and maximum accuracies at all feature combinations, and achieves great improvement over the importance analysis(IA)method on the whole. Moreover, the IIA method achieves the same highest accuracy as the one by the exhaustive method, and further improves the performance of question classification.
针对基于机器学习的问题分类中问句特征的组合, 提出了一种基于重要性和抑制性分析(importance-inhibition analysis, IIA)的特征组合方法.该方法在组合问句特征时不仅考虑了单个特征本身的重要性, 还考虑了待组合特征之间的抑制性.在中文问题集上的实验结果表明, IIA方法在所有的特征组合上都获得了平均精度和最高精度的提升, 总体上比单纯基于重要性分析(importance analysis, IA)的特征组合方法要更加高效;同时, IIA方法还获得了与穷举式特征组合方法同样的最高精度, 进一步提升了当前中文问题分类的性能.

References:

[1] Zhang Z C, Zhang Y, Liu T, et al. Advances in open-domain question answering [J]. Acta Electronica Sinica, 2009, 37(5):1058-1069.(in Chinese)
[2] Li X, Roth D. Learning question classifiers[C]//Proc of the 19th International Conference on Computational Linguistics. Taipei, China, 2002: 1-7.
[3] Li X, Roth D. Learning question classifiers: the role of semantic information[J]. Journal of Natural Language Engineering, 2006, 12(3): 229-250.
[4] Zhang D, Lee W. Question classification using support vector machines[C]//Proc of the 26th Annual International ACM SIGIR Conference on Research and Development in Information Retrieval. Toronto, Canada, 2003: 26-32.
[5] Huang Z H, Thint M, Qin Z C. Question classification using head words and their hypernyms[C]//Proc of the 2008 Conference on Empirical Methods in Natural Language Processing. Honolulu, Hawaii, USA, 2008: 927-936.
[6] Huang Z H, Thint M, Celikyilmaz A. Investigation of question classifier in question answering[C]//Proc of the 2009 Conference on Empirical Methods in Natural Language Processing. Singapore, 2009: 543-550.
[7] Li F T, Zhang X, Yuan J H, et al. Classifying what-type questions by head noun tagging[C]//Proc of the 22nd International Conference on Computational Linguistics. Manchester, UK, 2008: 481-488.
[8] Li X, Huang X J, Wu L D. Combined multiple classifiers based on TBL algorithm and their application in question classification [J]. Journal of Computer Research and Development, 2008, 45(3): 535-541.(in Chinese)
[9] Sun J G, Cai D F, Lu D X, et al. HowNet based Chinese question automatic classification [J]. Journal of Chinese Information Processing, 2007, 21(1):90-95.(in Chinese)
[10] Zhang Z C, Zhang Y, Liu T, et al. Chinese question classification based on identification of cue words and extension of training set [J]. Chinese High Technology Letters, 2009, 19(2): 111-118.(in Chinese)

Memo

Memo:
Biographies: Yang Sichun(1970—), male, graduate; Chen Jiajun(corresponding author), male, doctor, professor, chenjj@nju.edu.cn.
Foundation items: The National Natural Science Foundation of China(No.61003112, 61170181), the Open Research Fund of State Key Laboratory for Novel Software Technology of China(No.KFKT2010B02), the Key Project of Natural Science Research for Anhui Colleges of China(No.KJ2011A048).
Citation: Yang Sichun, Gao Chao, Yao Jiamin, et al. Feature combination via importance-inhibition analysis[J].Journal of Southeast University(English Edition), 2013, 29(1):22-26.[doi:10.3969/j.issn.1003-7985.2013.01.005]
Last Update: 2013-03-20