机器阅读理解的研究进展

doi:10.13190/j.jbupt.2019-111

北京邮电大学学报 ›› 2019, Vol. 42 ›› Issue (6): 1-9.doi: 10.13190/j.jbupt.2019-111

• 综述 • 下一篇

机器阅读理解的研究进展

王小捷, 白子薇, 李可, 袁彩霞

北京邮电大学计算机学院, 北京 100876

收稿日期:2019-06-11 发布日期:2019-11-15
通讯作者: 白子薇(1994-),女,博士生,E-mail:bestbzw@bupt.edu.cn. E-mail:bestbzw@bupt.edu.cn
作者简介:王小捷(1969-),男,教授.
基金资助:
国家自然科学基金项目（61906018）

Survey on Machine Reading Comprehension

WANG Xiao-jie, BAI Zi-wei, LI Ke, YUAN Cai-xia

School of Computer Science, Beijing University of Posts and Telecommunications, Beijing 100876, China

Received:2019-06-11 Published:2019-11-15

摘要/Abstract

摘要： 为便于厘清机器阅读理解任务的研究现状，按照答案来源，将机器阅读理解分为完形填空、片段选择、多项选择和答案生成4类.在统一的编码器-交互与推理-输出框架下对此4类任务的已有研究进行了综述，并描述了2种对此框架的可能扩展；最后讨论了机器阅读理解未来需要解决的问题.

关键词: 机器阅读理解, 编码器, 交互, 注意力机制

Abstract: In order to make clear the recent work of machine reading comprehension (MRC) tasks, they are divided into four types of subtasks with different sources of answers. They are cloze-style, span selection, multi-choice, and answer generation. Previous work on these four different types of subtasks is investigated under a unified framework of encoder-interaction and reasoning-output. Two types of recent developments for the frame are also given. Several challenges on MRC for future work are discussed at the end of the survey.

Key words: machine reading comprehension, encoder, interaction, attention mechanism

中图分类号:

TP18

王小捷, 白子薇, 李可, 袁彩霞. 机器阅读理解的研究进展[J]. 北京邮电大学学报, 2019, 42(6): 1-9.

WANG Xiao-jie, BAI Zi-wei, LI Ke, YUAN Cai-xia. Survey on Machine Reading Comprehension[J]. Journal of Beijing University of Posts and Telecommunications, 2019, 42(6): 1-9.

参考文献

[1] Wang Z, Liu J, Xiao X, et al. Joint training of candidate extraction and answer selection for reading comprehension[C]//ACL 2018. Melbourne:ACL, 2018:1715-1724.
[2] Heath J, Kamp H, Reyle U. From discourse to logic:introduction to model theoretic semantics of natural language, formal logic and discourse representation theory[J]. Language, 1995, 71(4):823.
[3] Hermann K M, Kocisky T, Grefenstette E, et al. Teaching machines to read and comprehend[C]//NIPS 2015. Montréal:MIT Press, 2015:1693-1701.
[4] Lai G, Xie Q, Liu H, et al. Race:large-scale reading comprehension dataset from examinations[C]//EMNLP 2017. Copenhagen:ACL, 2017:785-794.
[5] Rajpurkar P, Zhang J, Lopyrev K, et al. Squad:100,000+ questions for machine comprehension of text[C]//EMNLP 2016. Austin:ACL, 2016:2383-2392.
[6] Devlin J, Chang M W, Lee K, et al. Bert:pre-training of deep bidirectional transformers for language understanding[C]//NAACL-HIT 2019. Minneapolis:ACL, 2019:4171-4186.
[7] Trischler A, Wang T, Yuan X, et al. Newsqa:a machine comprehension dataset[C]//ACL workshop 2017. Vancouver:ACL, 2017:191-200.
[8] Nguyen T, Rosenberg M, Song X, et al. Ms-marco:a human generated machine reading comprehension dataset[EB/OL]. (2018-10-31)[2019-05-28]. https://www.microsoft.com/en-us/research/wp-content/uploads/2017/05/r-net.pdf.
[9] He W, Liu K, Liu J, et al. Dureader:a chinese machine reading comprehension dataset from real-world applications[C]//ACL workshop 2018. Melbourne:ACL, 2018:37-46.
[10] Lin C. Rouge:a package for automatic evaluation of summaries[C]//WAS 2004. Barcelona:ACL, 2004:74-81.
[11] Hirschman L, Light M, Breck E, et al. Deep read:a reading comprehension system[C]//ACL 1999. Maryland:ACL, 1999:325-332.
[12] Riloff E, Thelen M. A rule-based question answering system for reading comprehension tests[C]//NAACL/ANLP workshop 2000. ACL, 2000:13-19.
[13] Hao X, Chang X, Liu K. A rule-based chinese question answering system for reading comprehension tests[C]//IIH-MSP 2007.[S.l.]:IEEE, 2007:325-329.
[14] Cui Y, Chen Z, Wei S, et al. Attention-over-attention neural networks for reading comprehension[C]//ACL 2017. Vancouver:ACL, 2017:593-602.
[15] Wang S, Jiang J. Machine comprehension using match-lstm and answer pointer[C]//ICLR 2017. Toulon:Proceedings, 2017.
[16] Kadlec R, Schmid M, Bajgar O, et al. Text understanding with the attention sum reader network[C]//ACL 2016. Berlin:ACL, 2016:908-918.
[17] Chen D, Bolton J, Manning C D. A thorough examination of the cnn/daily mail reading comprehension task[C]//ACL 2016. Berlin:ACL, 2016:2358-2367.
[18] Cui Y, Liu T, Chen Z, et al. Consensus attention-based neural networks for chinese reading comprehension[C]//COLING 2016. Osaka:ACM Press, 2016:1777-1786.
[19] Sukhbaatar S, Weston J, Fergus R. End-to-end memory networks[C]//NIPS 2015. Montréal:MIT Press, 2015:2440-2448.
[20] Dhingra B, Liu H, Yang Z, et al. Gated-attention readers for text comprehension[C]//ACL 2017. Vancouver:ACL, 2017:1832-1846.
[21] Tseng B H, Shen S S, Lee H Y, et al. Towards machine comprehension of spoken content:initial toefl listening comprehension test by machine[C]//Interspeech 2016. San Francisco:ISCA, 2016:2731-2735.
[22] Zhu H, Wei F, Qin B, et al. Hierarchical attention flow for multiple-choice reading comprehension[C]//AAAI 2018. New Orleans:AAAI Press, 2018:6077-6085.
[23] Parikh S, Sai A B, Nema P, et al. Eliminet:a model for eliminating options for reading comprehension with multiple choice questions[C]//IJCAI 2018. Stockholm:Morgan Kaufmann, 2018:4272-4278.
[24] Xu Y, Liu J, Gao J, et al. Dynamic fusion networks for machine reading comprehension[EB/OL]. (2018-02-26)[2019-05-28]. https://arxiv.org/pdf/1711.04964.pdf.
[25] Seo M, Kembhavi A, Farhadi A, et al. Bidirectional attention flow for machine comprehension[C]//ICLR 2017. Toulon:Proceedings, 2017.
[26] Wang W, Yang N, Wei F, et al. R-net:machine reading comprehension with self-matching networks[EB/OL]. (2017-05-20)[2019-05-28]. https://www.microsoft.com/en-us/research/wp-content/uploads/2017/05/r-net.pdf.
[27] 桑志杰, 袁彩霞. 一种端到端的生成式问答模型[EB/OL]. 2019(2019-01-28)[2019-05-28]. http://www.paper.edu.cn/releasepaper/content/201901-186.
[28] Tan C, Wei F, Yang N, et al. S-net:from answer extraction to answer generation for machine reading comprehension[EB/OL]. 2018(2018-01-02)[2019-05-28]. http://arxiv.org/abs/1706.04815.
[29] See A, Liu P J, Manning C D. Get to the point:summarization with pointer-generator networks[C]//ACL 2017. Vancouver:ACL, 2017:1073-1083.
[30] Yu A W, Dohan D, Luong M T, et al. Qanet:combining local convolution with global self-attention for reading comprehension[C]//ICLR 2018. Vancouver:Proceedings, 2018.
[31] Nishida K, Saito I, Nishida K, et al. Multi-style generative reading comprehension[C]//ACL 2019, Florence:ACL, 2019:2273-2284.
[32] Zhuang Y, Wang H. Token-level dynamic self-attention network for multi-passage reading comprehension[C]//ACL 2019, Florence:ACL, 2019:2252-2262.
[33] Radford A, Narasimhan K, Salimans T, et al. Improving language understanding by generative pre-training[EB/OL]. 2018(2018-12-03)[2019-05-28]. https://www.cs.ubc.ca/~amuham01/LING530/papers/radford2018improving.pdf.
[34] Vaswani A, Shazeer N, Parmar N, et al. Attention is all you need[C]//NIPS 2017. Long Beach:MIT Press, 2017:5998-6008.
[35] Mihaylov T, Frank A. Knowledgeable reader:enhancing cloze-style reading comprehension with external commonsense knowledge[C]//ACL 2018, Melbourne:ACL, 2018:821-832.
[36] Hu M, Peng Y, Huang Z, et al. Retrieve, read, rerank:towards end-to-end multi-document reading comprehension[C]//ACL 2019, Florence:ACL, 2019:2285-2295.
[37] Wang Y, Liu K, Liu J, et al. Multi-passage machine reading comprehension with cross-passage answer verification[C]//ACL 2018, Melbourne:ACL, 2018:1918-1927.
[38] Xiao Y, Qu Y, Qiu L, et al. Dynamically fused graph network for multi-hop reasoning[C]//ACL 2019, Florence:ACL, 2019:6140-6150.
[39] Trischler A, Ye Z, Yuan X, et al. Natural language comprehension with the epireader[C]//EMNLP 2016, Austin:ACL, 2016:128-137.

机器阅读理解的研究进展

Survey on Machine Reading Comprehension

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

本文评价

[1]	包乾辉文娟石淑珍董萌萍刘雪. 融合句法依存与 BERT-Att-BiLSTM 的属性分类[J]. 北京邮电大学学报, 2023, 46(4): 123-128.
[2]	胡敏熊思黄宏程张光华王春芳. 基于RAPNet的脑肿瘤MRI图像三维分割[J]. 北京邮电大学学报, 2023, 46(2): 91-97.
[3]	朱军杨军李凯于文欣. 基于增强CVAE的三维射线重构方法[J]. 北京邮电大学学报, 2022, 45(5): 36-41.
[4]	刘琳岚, 宋修洋, 陈宇斌. 基于网络表示学习的机会网络链路预测[J]. 北京邮电大学学报, 2022, 45(4): 64-69,103.
[5]	宁静, 佘红艳, 赵东, 罗丹, 王磊. 一种路网级交通事故风险预测方法[J]. 北京邮电大学学报, 2022, 45(2): 72-78.
[6]	赵海英, 朱会, 侯小刚. 基于改进EMA单元的传统服饰图像语义分割[J]. 北京邮电大学学报, 2022, 45(1): 69-74.
[7]	王文竹, 肖波, 陈柯宏. 一种面向目标的情感极性分析方法[J]. 北京邮电大学学报, 2021, 44(5): 21-27.
[8]	霍纬纲, 王星, 梁锐. 融合互信息估计和对抗自编码器的异常检测[J]. 北京邮电大学学报, 2021, 44(5): 28-34.
[9]	李威, 陈曙东, 欧阳小叶, 杜蓉, 王荣. 基于高速多核网络的远监督关系抽取方法[J]. 北京邮电大学学报, 2020, 43(5): 71-76.
[10]	刘逸琛, 孙华志, 马春梅, 姜丽芬, 钟长鸿. 一种基于高层特征融合的网络商品分类[J]. 北京邮电大学学报, 2020, 43(5): 98-104,117.
[11]	戴庆达, 叶茂. 基于FPGA的高精度时间数字转换电路设计[J]. 北京邮电大学学报, 2020, 43(4): 88-94.
[12]	仇景明, 曲桦, 赵季红. 一种鲁棒网络流量分类及新类型发现算法[J]. 北京邮电大学学报, 2020, 43(2): 40-45.
[13]	黄毅, 冯俊兰, 胡珉, 吴晓婷, 杜晓宇. 智能对话系统架构及算法[J]. 北京邮电大学学报, 2019, 42(6): 10-19.
[14]	刘咏彬, 王小捷, 袁彩霞, 易炼. 通过检测语义分歧识别无答案问题[J]. 北京邮电大学学报, 2019, 42(6): 126-133,141.
[15]	朱皓, 谭咏梅. 基于胶囊的英文文本蕴含识别方法[J]. 北京邮电大学学报, 2019, 42(3): 21-28.