[1]姚 奕,尹瑞江,陈朝阳.问答系统构建及推理研究综述[J].计算机技术与发展,2023,33(12):8-16.[doi:10. 3969 / j. issn. 1673-629X. 2023. 12. 002]
　YAO Yi,YIN Rui-jiang,CHEN Zhao-yang.A Review of Question Answering System Construction and Inference Research[J].,2023,33(12):8-16.[doi:10. 3969 / j. issn. 1673-629X. 2023. 12. 002]

点击复制

问答系统构建及推理研究综述()

分享到：

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:

33

期数:

2023年12期

页码:

8-16

栏目:

综述

出版日期:

2023-12-10

文章信息/Info

Title:

A Review of Question Answering System Construction and Inference Research

文章编号:

1673-629X(2023)12-0008-09

作者:

姚 奕; 尹瑞江; 陈朝阳

中国人民解放军陆军工程大学 指挥控制工程学院,江苏 南京 210000

Author(s):

YAO Yi; YIN Rui-jiang; CHEN Zhao-yang

School of Command and Control Engineering,PLA Army Engineering University,Nanjing 210000,China

关键词:

答系统; 语义解析; 信息检索; 问答推理; 深度学习

Keywords:

question-answering system; semantic parsing; information retrieval; question-answering reasoning; deep learning

分类号:

TP391. 1

DOI:

10. 3969 / j. issn. 1673-629X. 2023. 12. 002

摘要:

近年来对问答系统的研究提高了信息提取的质量,并取得了许多领域内的不错成果。 传统方法构建的问答系统满足不了如今的需求,因此结合深度学习模型构建问答系统以提高检索能力成为当前研究的主流。 而且面对越来越多限制条件的多跳问题,问答系统需要具备一定推理能力推导出更多的信息以准确的找到答案。 该文讨论了基于语义解析和基于信息检索的两种实现问答系统的方法,这两种方法都可以有效地处理单一约束的简单问题,而且结合深度学习模型,可以更好地解决多约束的复杂问题。 此外,针对在知识库中多次跳跃的问答,还讨论了基于图神经网络和强化学习等方法的问答推理技术,这些技术可以在知识库中进行多跳推理,补充问答中的缺失信息完成问答任务。 最后,对两种构建方法的优缺点进行了总结,并展望了未来问答系统的发展前景。

Abstract:

In recent years,the research on question answering system has improved the quality of information extraction,and achievedgood results in many fields. The question - answering?

system constructed by traditional methods cannot meet today ’s needs,so it hasbecome the mainstream of current research to build a question-answering system to improve retrieval ability by combining deep learningmodel. In addition,there are more and more multi-hop questions with more and more constraints. The question answering system needsto have certain reasoning ability to deduce more information to find the answer accurately. We discuss two methods of question answeringsystem based on semantic parsing and information retrieval. Both of these methods can deal with simple problems with single constraintseffectively,and combined with deep learning model,can solve complex problems with multiple constraints better. In addition, for thequestion answering with multiple hops in the knowledge base,the question answering reasoning techniques based on graph neural networkand reinforcement learning are discussed. These techniques can perform multi - hop reasoning in the knowledge base and complete thequestion answering task by supplementing the missing information in the question answering. Finally,we summarize the advantages anddisadvantages of the two methods,and look forward to the future development of question answering system.

常用功能