期刊文献+

关于命名实体识别的生成式对抗网络的研究

Research on Generative Adversarial Networks of Named Entity Recognition
收藏 分享 导出
摘要 本文结合条件生成式对抗网络(CGAN)和改进的Wasserstein生成式对抗网络(WGAN-GP),提出一种适合于命名实体识别任务的条件Wasserstein生成式对抗网络模型(CWGAN).该模型借鉴CGAN以文本描述为条件的图像概率分布的思想,来完成以句子序列为条件获得标注序列概率分布的任务.该模型的生成器和判别器都采用BiLSTM结构,不同的是生成器生成命名实体标签的概率分布,判别器则为生成器的生成质量打分并反馈给生成器,生成器根据反馈更新梯度从而提升生成标签概率的质量.另外,CWGAN采用梯度惩罚的方法来保证梯度在反向传播的过程中保持平稳,通过拉近真实样本分布和生成样本之间的Wasserstein距离,优化目标函数.最后通过实验验证了该方法的可行性和优越性. This paper proposed a Generative Adversarial Nets suitable for the task of named entity recognition named Conditional Wasserstein Generative Adversarial Nets( CWGAN),inspired from Conditional GAN and improved Wasserstein GAN. Relative to the image probability distribution conditioned on textual description in CGAN,CWGAN obtains the NER label sequence probability distribution conditioned on sentence sequences. Both the generator and the discriminator use a bidirectional LSTM network. The difference is that the generator generates the probability distribution of the named entity tags,and the discriminator scores the generation quality of the generator and feeds it back to the generator. The generator updates the gradient according to the feedback to improve the quality of the probability of generating tags. In addition,this paper use gradient penalty in improved Wasserstein GAN to ensure that the gradient remains stable during backward propagation. Meanwhile,this paper use the mean which decrease the Wasserstein distance between real sample distribution and generate sample ensure that the target function is optimized. Experiments show that the CWGAN model we proposed is effective in the task of named entity recognition. Finally,the feasibility and superiority of the method are verified by experiments.
作者 冯建周 马祥聪 刘亚坤 宋沙沙 FENG Jian-zhou;MA Xiang-cong;LIU Ya-kun;SONG Sha-sha(Yanshan University College of Information Science and Engineering,Qinhuangdao 066004,China)(Yanshan University Key Laboratory of Hebei Software Engineering,Qinhuangdao 066004,China)
出处 《小型微型计算机系统》 CSCD 北大核心 2019年第6期1191-1196,共6页 Journal of Chinese Computer Systems
基金 国家自然科学基金青年基金项目(61602401)资助 河北省高等学校科学技术研究青年基金项目(QN2018074)资助.
关键词 命名实体识别 生成式对抗网络 BiLSTM Wasserstein距离 CWGAN named entity recognition generative adversarial networks bidirectional LSTM wasserstein distance conditional wasserstein generative adversarial nets(CWGAN)
作者简介 通讯作者:冯建周,男,1978年生,博士,副教授,CCF高级会员,研究方向为知识图谱、语义web,E-mail:fjzwxh@ysu.edu.cn;马祥聪,男,1994年生,硕士,研究方向为知识库补全;刘亚坤,男,1997年生,研究方向为命名实体识别、实体关系抽取;宋沙沙,女,1992年生,硕士,研究方向为命名实体识别、实体关系抽取.
  • 相关文献

参考文献2

二级参考文献34

  • 1He TT, Xu C, Li J, Zhao JZ. Named entity relation extraction method dased on seed self-expansion. Compter Engineering, 2006,32(21): 183-184, 193 (in Chinese with English abstract). 被引量:1
  • 2Fan N, Cai WD, Zhao Y. Extraction of subjective relation in opinion sentences based on maximum entropy model. Compter Engineering, 2010,36(2):4-6 (in Chinese with English abstract). 被引量:1
  • 3Che WX, Liu T, Li S. Automatic entity relation extraction. Journal of Chinese Information Processing, 2005,19(2): 1-6 (in Chinese with English abstract). 被引量:1
  • 4Huang X, Zhu QM, Qian LH, Liu MM. Chinese entity relation extraction based on features combination. Microelectronics & Computer, 2010,27(4): 198-200, 204 (in Chinese with English abstract). 被引量:1
  • 5Liu KB, Li F, Liu L, Hart Y. Implementation of a kernel-based Chinese relation extraction system. Journal of Computer Research and Development, 2007,44(8): 1406-1411 (in Chinese with English abstract). 被引量:1
  • 6Culotta A, Sorensen J. Dependency tree kernel for relation extraction. In: Proc. of the 42nd Annual Meeting of the Association for Computational Linguistics (ACL2004). Barcelona, 2004. 423-429. [doi: 10.3115/1218955.1219009]. 被引量:1
  • 7Bunescu RC, Mooney RJ. A shortest path dependency kernel for relation extraction. In: Proc. of the Human Language Technology Conf. and Conf. on Empirical Methods in Natural Language Processing. Vancouver, 2005. 724-731. [doi: 10.3115/1220575. 1220666]. 被引量:1
  • 8Zhang M, Zhang J, Su J. Exploring syntactic features for relation extraction using a convolution tree kernel. In: Proc. of the Human Language Technology Conf. of the North American Chapter of the Association for Computational Linguistics. New York: Springer-Verlag, 2006. 288-295. [doi: 10.3115/1220835.1220872]. 被引量:1
  • 9Zhao J, Wang XL, Guan Y. Comparing feature combination with features fusion in Chinese named entity recognition. Journal of Computer Applications, 2005,25(11):2647-2649 (in Chinese with English abstract). 被引量:1
  • 10Jing HY, Florian R, Luo XQ, Zhang T, Ittycheriah A. How to get a Chinese name (entity): Segmentation and combination issues. In: Proc. of the Conf. on Empirical Methods in Natural Language Processing. Sapporo, 2003. 200-207. 被引量:1

共引文献52

投稿分析

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部 意见反馈
新型冠状病毒肺炎防控与诊疗专栏