基于答案辅助的半监督问题分类方法

J4 ›› 2015, Vol. 37 ›› Issue (12): 2352-2357.

基于答案辅助的半监督问题分类方法

张栋，李寿山，周国栋

(苏州大学计算机科学与技术学院，江苏苏州 215006)

收稿日期:2015-08-15 修回日期:2015-10-24 出版日期:2015-12-25 发布日期:2015-12-25
基金资助:
国家自然科学基金重点项目（61331011）；国家自然科学基金资助项目（61375073,61273320）

A classification method for semi-supervised question classification with answers

ZHANG Dong，LI Shoushan，ZHOU Guodong

(School of Computer Science & Technology,Soochow University,Suzhou 215006,China)

Received:2015-08-15 Revised:2015-10-24 Online:2015-12-25 Published:2015-12-25

摘要/Abstract

摘要：

问题分类旨在对问题的类型进行自动分类，该任务是问答系统研究的一项基本任务。提出了一种基于答案辅助的半监督问题分类方法。首先，将答案特征结合问题特征一起实现样本表示；然后，利用标签传播方法对已标注问题训练分类器，自动标注未标注问题的类别；最后，将初始标注的问题和自动标注的问题合并作为训练样本，利用最大熵模型对问题的测试文本进行分类。实验结果表明，本文提出的基于答案辅助的半监督分类方法能够充分利用未标注样本提升性能，明显优于其他的基准方法。

关键词: 问答系统, 问题分类, 答案辅助, 半监督分类, 标签传播

Abstract:

Question classification aims at classifying the types of questions automatically, and this is a basic task of the question answering system. We propose a classification method for semi-supervised questions with answers. Firstly, we combine answer features with question features to realize sample expressions. Then we train a question classifier on labeled questions using label propagation algorithm to annotate the category of unlabeled questions automatically. The questions of initial annotation and automatic annotation are merged with each other as training samples, and the maximum entropy model is adopted to classify the testing samples. Experimental results demonstrate that the classification method for semisupervised questions with answers in this paper can make full use of the unlabeled samples to improve the performance, and it outperforms other benchmark methods.

Key words: question answering system;question classification;answer aiding;semi-supervised classification;label propagation

张栋，李寿山，周国栋. 基于答案辅助的半监督问题分类方法[J]. J4, 2015, 37(12): 2352-2357.

ZHANG Dong，LI Shoushan，ZHOU Guodong. A classification method for semi-supervised question classification with answers [J]. J4, 2015, 37(12): 2352-2357.

参考文献

［1］Li Xin,Huang Xuanjing,Wu Lide.Combined multiple classifiers based on TBL algorithm and their application in question classification［J］.Journal of Computer Research and Development,2008,45（3）:535541．(in Chinese)
［2］Gao Chao. Study on feature selection in Chinese question classification［D］.Anhui:Anhui University of Technology,2011.(in Chinese)
［3］Li S,Huang C R,Zhou G,et al.Employing personal/impersonal views in supervised and semisupervised sentiment classification［C］∥Proc of the 48th Annual Meeting of the Association for Computational Linguistics.Association for Computational Linguistics,2010:414423.
［4］Li S,Huang L,Wang J,et al.Semistacking for semisupervised sentiment classification［C］∥Proc of ACL2015,2015:2731.
［5］Ray S K,Singh S,Joshi B P.A semantic approach for question classification using WordNet and Wikipedia［J］.Pattern Recognition Letters,2010,31（13）:19351943．
［6］Hui Z,Liu J,Ouyang L.Question classification based on an extended class sequential rule model［C］∥Proc of IJCNLP, 2011:938946．
［7］Mishra M,Kumar Mishra V,Sharma H R.Question classification using semantic,syntactic and lexical features［J］.International Journal of Web & Semantic Technology,2013,4（3）:3947．
［8］Yadav R,Mishra M,Bhilai S.Question classification using Nave Bayes machine learning approach［J］.International Journal of Engineering and Innovative Technology （IJEIT）,2013,2（8）:291294．
［9］Ligozat A L. Question classification transfer［C］∥Proc of ACL, 2013:429433．
［10］Tian Weidong,Gao Yanying,Zu Yongliang.Question classification based on selflearning rules and modified Bayes［J］.Application Research of Computers,2010,27（8）:28692871．(in Chinese)
［11］Liu Xiaoming,Fan Xiaozhong,Li Fangfang.A question classification method combining domain ontology and question focus［J］.Transactions of Beijing Institute of Technology,2012,32（5）:498502.(in Chinese)
［12］Zhang Wei,Chen Junjie.Method of information entropy and its application in Chinese question classification［J］.Computer Engineering and Applications,2013,49（10）:129131．(in Chinese)
［13］Liu L,Yu Z,Guo J,et al.Chinese question classification based on question property kernel［J］.International Journal of Machine Learning & Cybernetics,2014,5（5）:713720．
［14］Zhang Junli,Chang Yanli,Shi Wen.Overview on label propagation algorithm and applications［J］.Application Research of Computers,2013,30（1）:2125.(in Chinese)
［15］Li S,Xue Y,Wang Z,et al.Active learning for crossdomain sentiment classification［C］∥Proc of the 32rd International Joint Conference on Artificial Intelligence,2013:21272133．
［16］Gao Wei,Wang Zhongqing,Li Shoushan.Semisupervised sentiment classification with a ensemble strategy［J］.Journal of Chinese Information Processing,2013,27（3）:120126．(in Chinese)
附中文参考文献：
［1］李鑫,黄萱菁,吴立德.基于错误驱动算法组合分类器及其在问题分类中的应用［J］.计算机研究与发展,2008,45（3）:535541．
［2］高超.中文问题分类中特征选择研究［D］.安徽：安徽工业大学,2011．
［10］田卫东,高艳影,祖永亮.基于自学习规则和改进贝叶斯结合的问题分类［J］.计算机应用研究,2010,27（8）:28692871．
［11］刘小明,樊孝忠,李方方.一种结合本体和焦点的问题分类方法［J］.北京理工大学学报,2012,32（5）:498502．
［12］张巍,陈俊杰.信息熵方法及在中文问题分类中的应用［J］.计算机工程与应用，2013,49（10）：129131．
［14］张俊丽,常艳丽,师文.标签传播算法理论及其应用研究综述［J］.计算机应用研究,2013,30（1）:2125．
［16］高伟,王中卿,李寿山.基于集成学习的半监督情感分类方法研究［J］.中文信息学报,2013,27（3）:120126．

[1]	皇甫斐斐, 杨阳, 邓晓懿, . 基于节点影响力的理性节点标签传播算法[J]. 计算机工程与科学, 2022, 44(04): 713-722.
[2]	潘曙灿, 许青林. 融合特征向量中心性与标签熵的标签传播算法[J]. 计算机工程与科学, 2020, 42(08): 1489-1499.
[3]	董建亮，赵文涛，李方,黄心昊. 一种基于标签传播算法的关键链路探测方法[J]. 计算机工程与科学, 2017, 39(11): 2022-2027.
[4]	柔特. 藏文问答系统中问句的分类方法研究[J]. J4, 2015, 37(07): 1393-1398.
[5]	王一萍，孙明. 应用人工鱼群算法的重叠社区检测[J]. J4, 2013, 35(10): 131-136.
[6]	陈小宾葛新伟林鸿飞. 基于语义计算的公交移动问答系统[J]. J4, 2008, 30(10): 118-121.