基于改进图节点的图神经多跳阅读理解研究

doi:10.19678/j.issn.1000-3428.0059917

摘要/Abstract

摘要： ：多跳阅读理解需要基于问题在多个支撑文档中寻找相关信息并进行跳跃式推理来回答问题。针对当前多跳阅读理解模型中所存在的实体图内缺乏关键问题信息以及信息量冗余的问题，提出了一种基于改进图节点的图神经多跳阅读理解模型。首先，采用基于语法的实体提取方法提取实体，将提取到的实体基于问题关联实体构建实体图。然后，对实体图中的节点进行编码预处理，并通过门机制的图卷积网络模拟推理得到推理信息。最后，计算推理信息与问题信息的双向注意力并进行结果预测。通过在 WikiHop 数据集中测试实验，结果表明相对于其他模型在测试集上效果有所提升，证明了该方法的有效性。

Abstract: Multi-hop reading comprehension requires searching for relevant information in supporting documents to answer the question according to the question understanding and leaping reasoning. In the current multi-hop reading comprehension models, the entity graph lacks pivotal answering information but contains superabundant useless information. To address this problem, we propose a multi-hop reading comprehension model based on graph neural network with improved graph nodes. Firstly, we build the entity graph based on the question-associated entities which are extracted based on syntax. After the nodes in the entity graph are preprocessed by encoding, we use the gated graph convolution network to obtain the inference information. Finally, we calculate the bidirectional attention between reasoning and question information and predict the answer. The experiment in the WikiHop dataset shows that the results are better than those of other models, which proves the effectiveness of the method.

舒冲 , 欧阳智 , 杜逆索 , 何庆 , 魏琴. 基于改进图节点的图神经多跳阅读理解研究[J]. 计算机工程, doi: 10.19678/j.issn.1000-3428.0059917.

SHU Chong , OUYANG Zhi , DU Nisuo , HE Qing , WEI Qin . Research on multi-hop reading comprehension based on graph neural network with improved graph nodes[J]. Computer Engineering, doi: 10.19678/j.issn.1000-3428.0059917.

参考文献

[1] Seo M, Kembhavi A, Farhadi A, et al. Bidirectional attention flow for machine comprehension[C]// Proceedings of the 5th International Conference on Learning Representations. Toulon, France, 2017.
[2] Xiong C M, Victor Z, Richard S. Dynamic coattention networks for question answering[C]// Proceedings of the 5th International Conference on Learning Representations. Toulon, France, 2017
[3] Liu X, Shen Y, Duh K, et al. Stochastic answer networks for machine reading comprehension[C]// Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). Melbourne, Australia, 2018:1694–1704.
[4] Chen D Q, Fisch A, Weston J, et al. Reading Wikipedia to answer open-domain questions [C]// Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). Vancouver, Canada, 2017:1870-1879.
[5] Clark C, Gardner M. Simple and effective multi-paragraph reading comprehension[C]// Proceedings of the 56th Annual Meeting of the Association for Computational Linguistic(Volume 1: Long Papers), Melbourne, Australia, 2018:845–855.
[6] WAN Jing, GUO YaZhi. Machine reading comprehension based on multi-passage ranking[J]. Journal of Beijing University of Chemical Technology (Natural Science), 2019, 46(3):93-98. (in Chinese) 万静,郭雅志.基于多段落排序的机器阅读理解研究[J].北京化工大学学报(自然科学版), 2019, 46(03):93-98.
[7] WU Ruizhi, ZHU Dayong, WANG Chunyu, QIN Ke. Location Semantics Inference with Graph Convolutional Networks[J]. Journal of University of Electronic Science and Technology of China, 2020, 49(05):739-744. (in Chinese) 吴睿智,朱大勇,王春雨,秦科.基于图卷积神经网络的位置语义推断[J].电子科技大学学报,2020,49(05):739-744.
[8] XU Li, LI Jianhua. Biomedical named entity recognition with graph network based on syntactic dependency parsing[J/OL]. Journal of Computer Applications:1-7[2020-11-30]. http:// kns.cnki.net/kcms/detail/51.1307.tp.20200910.1 010.002.html. (in Chinese) 许力,李建华.基于句法依存分析的图网络生物医学命名实体识别 [J/OL]. 计算机应用:1-7[2020-11-30]. http://kns.cnki.net/kcms/de tail/51.1307.tp.20200910.1010.002.html.
[9] Dhingra B, Jin Q, Yang Z, et al. Neural models for reasoning over multiple mentions using coreference[C]// Proceedings of the 16th Human Language Technology: Conference of the North American Chapter of the Association of Computational Linguistics. New Orleans, Louisiana, USA, 2018 : 42–48.
[10] De Cao N, Aziz W, Titov I. Question Answering by Reasoning Across Documents with Graph Convolutional Networks[C]// Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers). 2019: 2306-2317.
[11] Song L F, Wang Z G, Yu M, et al. Exploring graph-structured passage representation for multi-hop reading comprehension with graph neural networks [J] . arXiv preprint arXiv: 1809.02040, 2018.
[12] Chen J, Lin S, Durrett G. Multi-hop question answering via reasoning chains[J]. arXiv preprint arXiv:1910.02610, 2019.
[13] Tu M, Wang G, Huang J, et al. Multi-hop reading comprehension across multiple documents by reasoning over heterogeneous graphs[C]// Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics . Florence, Italy, 2019 : 2704–2713.
[14] Cao Y, Fang M, Tao D. Bag: Bi-directional attention entity graph convolutional network for multi-hop reasoning question answering[C]// Proceedings of the 17th Human Language Technology: Conference of the North American Chapter of the Association of Computational Linguistics. Minneapolis, Minnesota, USA, 2019 : 357–362.
[15] Peters M E, Neumann M, Iyyer M, Gardner M, Clark C, Lee K, and Zettlemoyer L. Deep contextualized word representations[C]// Proceedings of the 16th Human Language Technology: Conference of the North American Chapter of the Association of Computational Linguistics. New Orleans, Louisiana, USA, 2018:2227–2237.
[16] Hewlett D, Lacoste A, Jones L, et al. Wikireading: A novel large-scale language understanding task over Wikipedia[C]// Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). Berlin, Germany, 2016:1535–1545.
[17] Tang Z, Shen Y, Ma X, et al. Multi-hop Reading Comprehension across Documents with Path-based Graph Convolutional Network[J]. arXiv preprint arXiv:2006.06478, 2020.
[18] Jiang Y, Joshi N, Chen Y C, et al. Explore, propose, and assemble: An interpretable model for multi-hop reading comprehension[C]// Proceedings of the 57th Annual Meeting of theAssociation for Computational Linguistics. Florence, Italy, 2019: 2714–2725.
[19] Zhong V, Xiong C M, Keskar N S, et al. Coarse-grain fine-grain coattention network for multi-evidence question answering[C]// Proceedings of the 7th International Conference on Learning Representations. New Orleans, Louisiana, USA, 2019.
[20] Zhuang Y M, Wang H D. Token-level dynamic self-attention network for multi-passage reading comprehension[C]// Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics. Florence, Italy, 2019: 2252-2262.