您的位置：山东大学 -> 科技期刊社 -> 《山东大学学报（工学版）》

山东大学学报 (工学版) ›› 2020, Vol. 50 ›› Issue (3): 1-7.doi: 10.6040/j.issn.1672-3961.0.2019.417

• 机器学习与数据挖掘 •    下一篇

自然语言问答中的语义关系识别

段江丽1,2(),胡新2,*()   

  1. 1. 重庆邮电大学计算智能重庆市重点实验室, 重庆 400065
    2. 长江师范学院大数据与智能工程学院, 重庆 408100
  • 收稿日期:2019-07-22 出版日期:2020-06-20 发布日期:2020-06-16
  • 通讯作者: 胡新 E-mail:d180201004@stu.cqupt.edu.cn;huxin@yznu.edu.cn
  • 作者简介:段江丽(1989—)，女，云南怒江人，博士研究生，主要研究方向为粒计算，知识图谱，问答，数据挖掘. E-mail: d180201004@stu.cqupt.edu.cn
  • 基金资助:
    重庆邮电大学博士研究生人才培养项目(BYJS201908);重庆市教委科技研究计划青年项目(KJQN201901414);长江师范学院高层次人才科研启动金项目(0107/011160052)

Semantic relation recognition for natural language question answering

Jiangli DUAN1,2(),Xin HU2,*()   

  1. 1. Chongqing Key Laboratory of Computational Intelligence, Chongqing University of Posts and Telecommunications, Chongqing 400065, China
    2. College of Big Data and Intelligent Engineering, Yangtze Normal University, Chongqing 408100, China
  • Received:2019-07-22 Online:2020-06-20 Published:2020-06-16
  • Contact: Xin HU E-mail:d180201004@stu.cqupt.edu.cn;huxin@yznu.edu.cn
  • Supported by:
    重庆邮电大学博士研究生人才培养项目(BYJS201908);重庆市教委科技研究计划青年项目(KJQN201901414);长江师范学院高层次人才科研启动金项目(0107/011160052)

RichHTML

41

PDF (PC)

1113

摘要：

为了避免问句理解阶段过度依赖命名实体,通过语义关系理解中文自然语言问句中关键信息的逻辑关系,提出基于依赖结构的语义关系识别方法,从问句的依赖结构集中识别出对生成语义关系有价值的三类依赖结构集,将三类依赖结构集组合或转换得到语义关系。在中文标准问答数据集上的试验结果验证了本语义关系识别方法的有效性和可扩展性,本方法在命名实体识别失败时也可以理解中文自然语言问句。

关键词: 知识图谱, 自然语言问答, 语义关系, 依赖结构

Abstract:

To avoid the deficiency of excessive dependence on named entity recognition during the understanding perio, logic relationships among vital information in Chinese natural language question were understood by semantic relation. An algorithm for recognizing semantic relations based on dependency structures was proposed, which first recognized three kinds of valuable dependency structures that were vital for obtaining semantic relations, and then combined or transformed these dependency structures to obtain semantic relations. The effectiveness and scalability of the proposed method were verified by extensive experiments over Chinese benchmark question answering datasets, and the experiments results showed that this method could also understand Chinese natural language questions when recognition of named entity failed.

Key words: knowledge graph, natural language question answering, semantic relation, dependency structure

中图分类号: 

  • TP391

图1

依赖结构集和语义关系"

图2

依赖结构集和命名实体"

表1

三类依赖结构集"

分类 依赖结构集 备注
δsubject-predicate nsubj, nsubjpass, xsubj 主-谓依赖结构
δpredicate-object dobj 谓-宾依赖结构
δnmod nmod 谓-宾或主-宾依赖结构

表2

连接主-谓、谓-宾以及nmod的语义关系识别结果"

数据集 视角 Precision Recall F-measure
NLPCC2018 宏观 0.65 0.52 0.58
微观 0.45 0.52 0.47
CCKS2018 宏观 0.67 0.60 0.62
微观 0.58 0.57 0.58

表3

问句中的语义关系识别结果对比"

数据集 问句数量 正确识别语义关系的问句数量
本研究方法 其他方法[1-11]
NLPCC2018 122 61 0
CCKS2018 86 48 0

表4

命名实体识别为空的问句集上的语义关系识别结果"

数据集 视角 Precision Recall F-measure
NLPCC2018(61/122) 宏观 0.62 0.52 0.55
微观 0.45 0.50 0.47
CCKS2018(48/86) 宏观 0.62 0.49 0.58
微观 0.58 0.60 0.55

表5

仅连接主-谓和谓-宾的语义关系识别结果"

数据集 视角 Precision Recall F-measure
NLPCC2018 宏观 0.85 0.22 0.35
微观 0.22 0.22 0.22
CCKS2018 宏观 0.67 0.18 0.28
微观 0.17 0.18 0.17

图3

问句数量变化时的Precision"

图4

问句数量变化时的Recall"

图5

问句数量变化时的F-measure"

1 ZHANG J, HUA Y, QI G, et al. Semantic parsing for multiple-relation Chinese question answering[C]//Proc of the China Conf on Knowledge Graph and Semantic Computing (CCKS). Tianjin, China: CEUR-WS, 2018: 101-106.
2 LI Y, MIAO Q, YIN C, et al. A joint model of entity linking and predicate recognition for knowledge base question answering[C]// Proc of the China Conf on Knowledge Graph and Semantic Computing (CCKS). Tianjin, China: CEUR-WS, 2018: 95-100.
3 SUN Z, SONG L, YU J. A QA search algorithm based on the fusion integration of text similarity and graph computation[C]//Proc of the China Conf on Knowledge Graph and Semantic Computing (CCKS). Tianjin, China: CEUR-WS, 2018: 89-94.
4 NI H, LIN L, XU G. A relateness-based ranking method for knowledge-based question answering[C]//Proc of the Seventh CCF Int Conf on Natural Language Processing and Chinese Computing (NLPCC). Huhehot, China: Springer, 2018: 393-400.
5 LAI Y, JIA Y, LIN Y, et al. A Chinese question answering system for single-relation factoid questions[C]//Proc of the Sixth CCF Int Conf on Natural Language Processing and Chinese Computing (NLPCC). Dalian, China: Springer, 2017: 124-135.
6 LAI Y, LIN Y, CHEN J, et al. Open domain question answering system based on knowledge base[C]//Proc of the Natural Language Processing and Chinese Computing (NLPCC). Kunming, China: Springer, 2016: 722-733.
7 LEI K, DENG Y, ZHANG B, et al. Open domain question answering with character-level deep learning models[C]//Proc of the 10th Int Symposium on Computational Intelligence and Design (ISCID). Hangzhou, China: IEEE, 2017: 30-33.
8 SHEN C , HUANG T , LIANG X , et al. Chinese knowledge base question answering by attention-based multi-granularity model[J]. Information, 2018, 9 (4): 1- 20.
9 YANG F, GAN L, LI A, et al. Combining deep learning with information retrieval for question answering[C]//Proc of the Natural Language Processing and Chinese Computing (NLPCC). Kunming, China: Springer, 2016: 917-925.
10 XIE Z, ZENG Z, ZHOU G, et al. Knowledge base question answering based on deep learning models[C]// Proc of the Natural Language Processing and Chinese Computing (NLPCC). Kunming, China: Springer, 2016: 300-311.
11 WANG L, ZHANG Y, LIU T. A deep learning approach for question answering over knowledge base[C]// Proc of the Natural Language Processing and Chinese Computing (NLPCC). Kunming, China: Springer, 2016: 885-892.
12 PICHUAN C A, HUIHSIN T B, DAN J A, et al. Discriminative reordering with chinese grammatical relations features[C]//Proc of the Third Workshop on Syntax and Structure in Statistical Translation. Pennsylvania, USA: Association for Computational Linguistics, 2009: 51-59.
13 DE MARNEFFE M C, DOZAT T, SILVEIRA N, et al. Universal Stanford dependencies: A cross-linguistic typology[C]// Proc of the International Conference on Language Resources and Evaluation (LREC). Reykjavik, Iceland: European Language Resources Association, 2014: 4585-4592.
14 YAHYA M, BERBERICH K, ELBASSUONI S, et al. Deep answers for naturally asked questions on the web of data[C]//Proc of the 21st Int Conf on World Wide Web(WWW). Lyon, France: ACM, 2012: 445-449.
15 YAHYA M, BERBERICH K, ELBASSUONI S, et al. Robust question answering over the web of linked data[C]//Proc of the 22nd ACM Int Conf on Information & Knowledge Management (CIKM). New York, USA: Association for Computing Machinery, 2013: 1107-1116.
16 YAHYA M, BERBERICH K, ELBASSUONI S, et al. Natural language questions for the web of data[C]//Proc of the 2012 Joint Conf on Empirical Methods in Natural Language Processing and Computational Natural Language Learning. Jeju Island, Korea: Association for Computational Linguistics, 2012: 379-390.
17 YAHYA M. Question answering and query processing for extended knowledge graphs[D]. Saarland, German: Saarland University, 2016.
18 PRADEL C, PEYET G, HAEMMERLÉ O, et al. SWIP at QALD-3: Results, criticisms and lesson learned[C]// Proc of the 3rd Open Challenge on Question Answering over Linked Data (QALD). Valencia, Spain: CEUR-WS, 2013: 1-13.
19 PRADEL C, HAEMMERLÉ O, HERNANDEZ N. Natural language query translation into SPARQL using patterns[C]//Proc of the Fourth Int Workshop on Consuming Linked Data. Sydney, Australia: CEUR-WS, 2013: 1-12.
20 PRADEL C, HAEMMERLÉ O, HERNANDEZ N. SWIP: a natural language to SPARQL interface implemented with SPARQL[C]//Proc of the Int Conf on Conceptual Structures. Işi, Romania: Springer, 2014: 260-274.
21 XU K, FENG Y, HUANG S, et al. Question answering via phrasal semantic parsing[C]//Proc of the Int Conf of the Cross-Language Evaluation Forum for European Languages. Toulouse, France: Springer, 2015: 414-426.
22 XU K, ZHANG S, FENG Y, et al. Answering natural language questions via phrasal semantic parsing[C]//Proc of the Natural Language Processing and Chinese Computing (NLPCC). Shenzhen, China: Springer, 2014: 333-344.
23 HAKIMOV S, TUNC H, AKIMALIEV M, et al. Semantic question answering system over linked data using relational patterns[C]// Proc of the Joint EDBT/ICDT 2013 Workshops. Genoa, Italy: ACM, 2013: 83-88.
24 ZOU L, HUANG R, WANG H, et al. Natural language question answering over RDF: a graph data driven approach[C]//Proc of the 2014 ACM SIGMOD Int Conf on Management of Data. Utah, USA: ACM, 2014: 313-324.
25 LIU J , LI W , LUO L , et al. Linked open data query based on natural language[J]. Chinese Journal of Electronics, 2017, 26 (2): 230- 235.
doi: 10.1049/cje.2016.11.003
26 FINKEL J R, GRENAGER T, MANNING C. Incorporating non-local information into information extraction systems by gibbs sampling[C]//Proc of the 43rd Annual Meeting of the Association for Computational Linguistics (ACL). Michigan, USA: Association for Computational Linguistics, 2005: 363-370.
[1] 田轶群,林荣恒. 基于知识图谱的查询显示系统的设计与实现[J]. 山东大学学报 (工学版), 2022, 52(2): 67-73.
[2] 龚乐君,杨璐,高志宏,李华康. LncRNA与疾病关系的知识图谱构建[J]. 山东大学学报 (工学版), 2021, 51(2): 26-33.
[3] 苏佳林,王元卓,靳小龙,程学旗. 自适应属性选择的实体对齐方法[J]. 山东大学学报 (工学版), 2020, 50(1): 14-20.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
[1] 王素玉,艾兴,赵军,李作丽,刘增文 . 高速立铣3Cr2Mo模具钢切削力建模及预测[J]. 山东大学学报(工学版), 2006, 36(1): 1 -5 .
[2] 李 侃 . 嵌入式相贯线焊接控制系统开发与实现[J]. 山东大学学报(工学版), 2008, 38(4): 37 -41 .
[3] 施来顺,万忠义 . 新型甜菜碱型沥青乳化剂的合成与性能测试[J]. 山东大学学报(工学版), 2008, 38(4): 112 -115 .
[4] 孔祥臻,刘延俊,王勇,赵秀华 . 气动比例阀的死区补偿与仿真[J]. 山东大学学报(工学版), 2006, 36(1): 99 -102 .
[5] 陈瑞,李红伟,田靖. 磁极数对径向磁轴承承载力的影响[J]. 山东大学学报(工学版), 2018, 48(2): 81 -85 .
[6] 李可,刘常春,李同磊 . 一种改进的最大互信息医学图像配准算法[J]. 山东大学学报(工学版), 2006, 36(2): 107 -110 .
[7] 季涛,高旭,孙同景,薛永端,徐丙垠 . 铁路10 kV自闭/贯通线路故障行波特征分析[J]. 山东大学学报(工学版), 2006, 36(2): 111 -116 .
[8] 浦剑1 ,张军平1 ，黄华2 . 超分辨率算法研究综述[J]. 山东大学学报(工学版), 2009, 39(1): 27 -32 .
[9] 王丽君,黄奇成,王兆旭 . 敏感性问题中的均方误差与模型比较[J]. 山东大学学报(工学版), 2006, 36(6): 51 -56 .
[10] 孙殿柱,朱昌志,李延瑞 . 散乱点云边界特征快速提取算法[J]. 山东大学学报(工学版), 2009, 39(1): 84 -86 .
版权所有 © 2018 《山东大学学报(工学版)》编辑部　
地址: 济南市山大南路27号（250100）　 电话: 0531-88366735　 E-mail: xbgxb@sdu.edu.cn
本系统由北京玛格泰克科技发展有限公司设计开发