结合实体描述信息的跨句包关系抽取方法

doi:10.19678/j.issn.1000-3428.0058189

摘要/Abstract

摘要： 远程监督关系抽取方法能够大幅减少标注成本，但现有方法忽略了关系间的关联信息和实体背景知识。结合实体描述信息提出一种新的跨句包关系抽取方法。引入分段卷积神经网络进行句编码，解决特征提取的误差传播问题。同时设计跨关系跨句包注意力机制获取关系特征，更好地从远程监督的噪声数据中鉴别有效实例，从而充分利用关系之间丰富的相关信息并降低噪音句子的影响。在此基础上，利用卷积神经网络提取实体描述信息，补充关系抽取任务所需的背景知识，为跨关系跨句包注意力模块提供更好的实体表示。在NYT公共数据集上的实验结果表明，该方法在句子层面抽取任务上的F1值较结合句注意力与实体描述信息的分段卷积方法提高了4%左右，能够有效改善远程监督关系抽取效果。

关键词: 关系抽取, 实体描述, 跨关系注意力, 跨句包注意力, 远程监督

Abstract: Distant supervision can significantly reduce the cost of labeling, but the existing methods ignore the correlation information between relations and entity description information.To address the problem, this paper proposes a new cross-sentence bag relation extraction method combining entity description information.The Piecewise Convolutional Neural Network(PCNN) is introduced to encode sentences to alleviate the error propagation problem of feature extraction.At the same time, a cross-relation and cross-sentence bag attention mechanism is designed to obtain the relation features, so the valid instances can be identified more effectively from the noisy data of distant supervision. Thus the abundant relevant information between relations can be made full use of, and the influence of noisy sentences can be reduced.On this basis, CNN is used to extract entity description information, which supplements the background knowledge required by the relation extraction tasks and provides a better entity representation for the cross-relation and cross-sentence bag attention mechanism.The experimental results on NYT corpus show that the F1 score of the proposed method is improved by about 4% in sentence-level extraction tasks compared with the piecewise convolutional method combining sentence attention and entity description information.It can significantly improve the performance of relation extraction using distant supervision.

Key words: relation extraction, entity description, cross-relation attention, cross-sentence bag attention, distant supervision

中图分类号:

TP18

孙新, 申长虹, 姜景虎, 崔家铭. 结合实体描述信息的跨句包关系抽取方法[J]. 计算机工程, 2021, 47(6): 68-75.

SUN Xin, SHEN Changhong, JIANG Jinghu, CUI Jiaming. Cross-Sentence Bag Relation Extraction Method Combining Entity Description Information[J]. Computer Engineering, 2021, 47(6): 68-75.

http://www.ecice06.com/CN/Y2021/V47/I6/68

图/表 6

20210618124220

20210618124223

20210618124227

20210618124231

20210618124234

20210618124238

参考文献

[1] MINTZ M,BILLS S,SNOW R,et al.Distant supervision for relation extraction without labeled data[C]//Proceedings of the Joint Conference of the 47th Annual Meeting of the ACL and the 4th International Joint Conference on Natural Language Processing of the AFNLP.New York,USA:ACM Press,2009:1003-1011.
[2] HOFFMANN R,ZHANG C L,LING X,et al.Knowledge-based weak supervision for information extraction of overlapping relations[C]//Proceedings of the 49th Annual Meeting of the Association for Computational Linguistics:Human Language Technologies.[S.l.]:ACL,2011:541-550.
[3] SURDEANU M,TIBSHIRANI J,NALLAPATI R,et al.Multi-instance multi-label learning for relation extraction[C]//Proceedings of 2012 Joint Conference on Empirical Methods in Natural Language Processing and Computational Natural Language Learning.New York,USA:ACM Press,2012:455-465.
[4] ZENG Daojian,LIU Kang,CHEN Yubo,et al.Distant supervision for relation extraction via piecewise convolutional neural networks[C]//Proceedings of 2015 Conference on Empirical Methods in Natural Language Processing.[S.l.]:ACL,2015:1753-1762.
[5] LIN Yankai,SHEN Shiqi,LIU Zhiyuan,et al.Neural relation extraction with selective attention over instances[C]//Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics.[S.l.]:ACL,2016:2124-2133.
[6] LIU Tianyu,WANG Kexiang,CHANG Baobao,et al.A soft-label method for noise-tolerant distantly supervised relation extraction[C]//Proceedings of 2017 Conference on Empirical Methods in Natural Language Processing.[S.l.]:ACL,2017:1790-1795.
[7] HAN Xu,LIU Zhiyuan,SUN Maosong.Neural knowledge acquisition via mutual attention between knowledge graph and text[C]//Proceedings of AAAI Conference on Artificial Intelligence.[S.l.]:AAAI Press,2018:4832-4839.
[8] LI Lingfeng,NIE Yuanping,HAN Weihong,et al.A multi-attention-based bidirectional long short-term memory network for relation extraction[C]//Proceedings of International Conference on Neural Information Processing.Berlin,Germany:Springer,2017:216-227.
[9] E Haihong,ZHANG Wenjing,XIAO Siqi,et al.Survey of entity relationship extraction based on deep learning[J].Journal of Software,2019,30(6):1793-1818.(in Chinese)鄂海红,张文静,肖思琪,等.深度学习实体关系抽取研究综述[J].软件学报,2019,30(6):1793-1818.
[10] SUN Ziyang,GU Junzhong,YANG Jing.Chinese entity relation extraction method based on deep learning[J].Computer Engineering,2018,44(9):164-170.(in Chinese)孙紫阳,顾君忠,杨静.基于深度学习的中文实体关系抽取方法[J].计算机工程,2018,44(9):164-170.
[11] MILLER S,FOX H,RAMSHAW L,et al.A novel use of statistical parsing to extract information from text[C]//Proceedings of the 1st North American Chapter of the Association for Computational Linguistics Conference.[S.l.]:ACL,2000:226-233.
[12] JAYRAM T S,KRISHNAMURTHY R,RAGHAVAN S,et al.Avatar information extraction system[J].IEEE Data Engineering Bulletin,2006,29(1):40-48.
[13] JIANG Jing,ZHAI Chengxiang.A systematic exploration of the feature space for relation extraction[C]//Proceedings of Human Language Technology Conference of the North American Chapter of the Association of Computational Linguistics.[S.l.]:ACL,2007:113-120.
[14] MIAO Qingliang,ZHANG Shu,ZHANG Bo,et al.Extracting and visualizing semantic relationships from Chinese biomedical text[C]//Proceedings of the 26th Pacific Asia Conference on Language,Information and Computation.Bali,Indonesia:[s.n.],2012:99-107.
[15] BUNESCU R C,MOONEY R J.A shortest path dependency kernel for relation extraction[C]//Proceedings of Conference on Human Language Technology and Empirical Methods in Natural Language Processing.New York,USA:ACM Press,2005:724-731.
[16] ZHANG Min,ZHANG Jie,SU Jian,et al.A composite kernel to extract relations between entities with both flat and structured features[C]//Proceedings of the 21st International Conference on Computational Linguistics and the 44th Annual Meeting of the Association for Computational Linguistics.[S.l.]:ACL,2006:825-832.
[17] RIEDEL S,YAO L M,ANDREW M.Modeling relations and their mentions without labeled text[C]//Proceedings of Joint European Conference on Machine Learning and Knowledge Discovery in Databases.New York,USA:ACM Press,2010:148-163.
[18] ZENG Daojian,LIU Kang,LAI Siwei,et al.Relation classification via convolutional deep neural network[C]//Proceedings of the 25th International Conference on Computational Linguistics.Dublin,Ireland:[s.n.],2014:2335-2344.
[19] ZHOU Peng,SHI Wei,TIAN Jun,et al.Attention-based bidirectional long short-term memory networks for relation classification[C]//Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics.[S.l.]:ACL,2016:207-212.
[20] QIN P D,XU W R,WANG W Y.Robust distant supervision relation extraction via deep reinforcement learning[C]//Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics.[S.l.]:ACL,2018:2137-2147.
[21] FENG Jun,HUANG Minlie,ZHAO Li,et al.Reinforcement learning for relation classification from noisy data[C]//Proceedings of the 32nd AAAI Conference on Artificial Intelligence.[S.l.]:AAAI Press,2018:1-9.
[22] BORDES A,USUNIER N,GARCIA-DURAN A,et al.Translating embeddings for modeling multi-relational data[C]//Proceedings of the 26th International Conference on Neural Information Processing Systems.New York,USA:ACM Press,2013:2787-2795.

选择文件类型/文献管理软件名称

选择包含的内容