基于BERT-DGCNN的中文事件抽取方法研究
Research on Chinese Event Extraction Method Based on BERT-DGCNN
DOI: 10.12677/CSA.2021.115162, PDF,  被引量    科研立项经费支持
作者: 陈安南, 叶岩宁, 王畅畅, 王文举, 李博文：合肥工业大学，安徽 宣城
关键词: 事件抽取；BERT模型；膨胀门卷积神经网络；Event Extraction； BERT Model； Dilated Gated CNN
摘要: 本文构建了一个事件抽取pipeline模型，其旨在对新闻中的信息元进行有效的抽取。在管道抽取模式下，先对文本进行存在事件类型识别，而后再将事件类型与文本一并作为输入传入模型进行事件论元角色抽取，其中事件论元角色采用类似于BERT中SQuAD等阅读理解任务上的双指针输出。两个基本模型都是利用BERT预训练模型产生的词嵌入，使用DGCNN进行编码之后池化，再连接到dense层进行分类。实验结果表明，本模型可对新闻类内容进行高效抽取。
Abstract: This paper constructs an event extraction model, which aims at effectively extracting information elements from the news. This model is based on the BERT pretraining model, which first identifies the existing event type of the text, and then inputs the event type and text into the model to extract the event role, in which the event role adopts double-pointer output similar to the reading comprehension task of a SQuAD in BERT. Experimental results show that this model can extract news content efficiently.
文章引用：陈安南, 叶岩宁, 王畅畅, 王文举, 李博文. 基于BERT-DGCNN的中文事件抽取方法研究[J]. 计算机科学与应用, 2021, 11(5): 1572-1578. https://doi.org/10.12677/CSA.2021.115162

参考文献

[1] 吴刚. 基于主题的中文事件抽取技术研究及应用[D[: [硕士学位论文]. 苏州: 苏州大学, 2009.
[2] 郑家恒, 王兴义, 李飞. 信息抽取模式自动生成方法的研究[J]. 中文信息学报, 2004, 18(1): 48-54.
[3] Yangarber, R. (2001) Scenario Customization for Information Extraction. Defense Advanced Research Projects Agency, Arlington, VA.
[4] 姜吉发. 自由文本的信息抽取模式获取的研究[D]: [博士学位论文]. 北京: 中国科学院计算技术研究所, 2004.
[5] Ahn, D. (2006) The Stages of Event Extraction. Proceedings of the Workshop on Annotating and Reasoning about Time and Events, Association for Computational Linguistics, 1-8. [Google Scholar] [CrossRef] 
[6] 赵妍妍. 中文事件抽取的相关技术研究[D[: [硕士学位论文]. 哈尔滨: 哈尔滨工业大学, 2007.
[7] Chen, Y., Xu, L., Liu, K., et al. (2015) Event Extraction via Dynamic Mul-ti-Pooling Convolutional Neural Networks. Proceedings of the 53rd Annual Meeting of the Association for Computation-al Linguistics and the 7th International Joint Conference on Natural Language Processing (Volume 1: Long Papers), Association for Computational Linguistics, 167-176. [Google Scholar] [CrossRef] 
[8] Nguyen, T.H., Cho, K. and Grishman, R. (2016) Joint Event Extraction via Recurrent Neural Networks. Proceedings of the 2016 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, As-sociation for Computational Linguistic, 300-309. [Google Scholar] [CrossRef] 
[9] Mikolov, T., Chen, K., Corrado, G., et al. (2013) Efficient Estimation of Word Representations in Vector Space. arXiv:1301.3781 [cs.CL]
[10] Peters, M., Neumann, M., Iyyer, M., et al. (2018) Deep Contextualized Word Repre-sentations. Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long Papers), Association for Computational Linguistics, 2227-2237. [Google Scholar] [CrossRef] 
[11] Radford, A., Narasimhan, K., Salimans, T. and Sutskever, I. (n.d.) Im-proving Language Understanding by Generative PreTraining.
https://www.cs.ubc.ca/~amuham01/LING530/papers/radford2018improving.pdf
[12] Devlin, J., Chang, M.W., Lee, K., et al. (2018) Bert: Pre-Training of Deep Bidirectional Transformers for Language Understanding. arXiv:1810.04805 [cs.CL]