计算机科学 ›› 2016, Vol. 43 ›› Issue (1): 94-97, 102.doi: 10.11896/j.issn.1002-137X.2016.01.022

• 第五届全国智能信息处理学术会议 • 上一篇    下一篇

多特征层次化答案质量评价方法研究

崔敏君,段利国,李爱萍   

  1. 太原理工大学计算机科学与技术学院 太原030024,太原理工大学计算机科学与技术学院 太原030024,太原理工大学计算机科学与技术学院 太原030024;武汉大学软件工程国家重点实验室 武汉430072
  • 出版日期:2018-12-01 发布日期:2018-12-01
  • 基金资助:
    本文受武汉大学软件工程国家重点实验室开放课题项目(SKLSE2012-09-30),山西省自然科学基金项目(2013011015-2)资助

Research on Multi-features Hierarchical Answer Quality Evaluation Method

CUI Min-jun, DUAN Li-guo and LI Ai-ping   

  • Online:2018-12-01 Published:2018-12-01

摘要: 社交媒体中的问答对可以为自动问答系统提供答案,但有些答案的质量不高,因此答案质量评价方法具有研究价值。已有的评价方法没有考虑问题类别特征,对不同类型的问题采用统一的评价方法。因此提出了一个层次分类模型。首先分析问题类型;然后提取文本、非文本、语言翻译性、答案中的链接数4类特征,依据特征分类影响力随问题类型不同而不同这一客观现象,采用逻辑回归算法对各类型问题的答案质量进行评价,取得了较好的实验效果;最后分析了影响各类问题答案质量的主要特征。

关键词: 层次分类模型,问题类别,答案质量评价,特征分析

Abstract: Social media question-answer pairs can provide the answer to the automatic question and answering system,but the quality about some of the answers is not so high.So the evaluation method of answer quality has the research value.The existing evaluation methods without considering the problem of question types feature use the uniform evaluation method for different question types.This paper presented a hierarchical classification model.Firstly we analyzed the types of question,and then extracted four features of text,non-text,language translation,number of links in the answer.According to this objective phenomenon that the influence of feature classification varies with the types of different questions,we used logistic regression algorithm to evaluate various types of answer quality based on these features,achieving good results.Finally the main features that influence the anawer quality of all kinds of questions were analyzed.

Key words: Hierarchical classification model,Question types,Answer quality evaluation,Feature analysis

[1] Berger A,Della Pietra S,Della Pietra V J.A maximum entropy approach to natural language processing[J].Computational Linguistics,1996,2(1):39-71
[2] Hwang J N,Lay S R,Lippman A.Nonparametric multivariate density estimation:A comparative study[J].IEEE Transactions of Signal Processing,1994,2(10):2795-2810
[3] Zhu Z,Bernhard D,Gurevych I.A Multi-Dimensional model for assessing the quality of answers in social Q&A sites[R] .Technische Universitt UKP Lab,2009
[4] Liu G J,Ma Y Z,Duan J Y.Assessment of Quality of “Questions and Answerw” in Community Q&A System[J].Journal of North China University of Technology,2012,4(3):31-35(in Chinese)刘高军,马砚忠,段建勇.社区问答系统中“问答对”的质量评价[J].北方工业大学学报,2012,4(3):31-35
[5] Tian Z H.Answer selection for non-factoid question[D].Harbin:Harbin Institute of Technology,2013(in Chinese)田作辉.非事实类问题的答案选取[D].哈尔滨:哈尔滨工业大学,2013
[6] Li S G,Manandhar S.Improving question recommendation byexploiting information need[C]∥Proceedings of the 49th Annual Meeting of the Association for Computational Lin-guistics.Portland,Oregon:ACL,2011:1425-1434
[7] Figueroa A,Neumann G.Category-specific models for ranking effective paraphrase in community Question Answering[J].Expert System With Applications,2014,1(10):4730-4742
[8] Wang B X,Liu B Q,Wang X L,et al.Deep learning approaches to semantic relevance modeling for Chinese question-answer pairs[J].ACM Transactions on Asian Language Information Processing,2011,10(4):1-16
[9] Li C,Chao W H,Chen X M.Quality evaluation and prediction for question and answer in chinese community question answe-ring[J].Computer Science,2011,38(6):230-236(in Chinese)李晨,巢文涵,陈小明.中文社区问答中问题答案质量评价和预测[J].计算机科学,2011,38(6):230-236
[10] Wen X,Zhang Y,Liu T,et al.Syntactic Structure Parsing Based Chinese Question Classification[J].Journal of Chinese Information Processing,2006,0(2):33-39(in Chinese)文勖,张宇,刘挺,等.基于句法结构分析的中文问题分类[J].中文信息学报,2006,0(2):33-39
[11] Hu H F.Research on the method of the feature representation and fusion for evaluating the quality of user generated answers[D].Harbin:Harbin Institute of Technology,2013(in Chinese)胡海峰.用户生成答案质量评价中的特征表示及融合研究[D].哈尔滨:哈尔滨工业大学,2013
[12] Kong W Z,Liu Y Q,Zhang M,et al.Answer quality analysis on community question answering[J].Journal of Chinese Information Processing,2011(1):3-8(in Chinese)孔维泽,刘奕群,张敏,等.问答社区中回答质量的评价方法研究[J].中文信息学报,2011(1):3-8
[13] Yang H T,Wang J,Ling H F.Question sentence similarity computing based on multi-features fusion in community question answering[J].Journal of Jiangxi Normal University,2013,37(2):125-129(in Chinese)杨海天,王健,林鸿飞.基于特征融合的社区问答问句相似度计算[J].江西师范大学学报,2013,37(2):125-129
[14] Zhou Z M.The study of models and features for non-factoid question answering[D].Shanghai:East China Normal University,2012(in Chinese)周志敏.非事实类问题问答模型和特征的研究[D].上海:华东师范大学,2012

No related articles found!
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
[1] 雷丽晖,王静. 可能性测度下的LTL模型检测并行化研究[J]. 计算机科学, 2018, 45(4): 71 -75, 88 .
[2] 夏庆勋,庄毅. 一种基于局部性原理的远程验证机制[J]. 计算机科学, 2018, 45(4): 148 -151, 162 .
[3] 厉柏伸,李领治,孙涌,朱艳琴. 基于伪梯度提升决策树的内网防御算法[J]. 计算机科学, 2018, 45(4): 157 -162 .
[4] 王欢,张云峰,张艳. 一种基于CFDs规则的修复序列快速判定方法[J]. 计算机科学, 2018, 45(3): 311 -316 .
[5] 孙启,金燕,何琨,徐凌轩. 用于求解混合车辆路径问题的混合进化算法[J]. 计算机科学, 2018, 45(4): 76 -82 .
[6] 张佳男,肖鸣宇. 带权混合支配问题的近似算法研究[J]. 计算机科学, 2018, 45(4): 83 -88 .
[7] 伍建辉,黄中祥,李武,吴健辉,彭鑫,张生. 城市道路建设时序决策的鲁棒优化[J]. 计算机科学, 2018, 45(4): 89 -93 .
[8] 刘琴. 计算机取证过程中基于约束的数据质量问题研究[J]. 计算机科学, 2018, 45(4): 169 -172 .
[9] 钟菲,杨斌. 基于主成分分析网络的车牌检测方法[J]. 计算机科学, 2018, 45(3): 268 -273 .
[10] 史雯隽,武继刚,罗裕春. 针对移动云计算任务迁移的快速高效调度算法[J]. 计算机科学, 2018, 45(4): 94 -99, 116 .