计算机科学 ›› 2021, Vol. 48 ›› Issue (10): 77-84.doi: 10.11896/jsjkx.210300271

• 人工智能* 上一篇    下一篇

基于深度神经网络和自注意力机制的医学实体关系抽取

张世豪, 杜圣东, 贾真, 李天瑞   

  1. 西南交通大学计算机与人工智能学院 成都611756
  • 收稿日期:2021-03-29 修回日期:2021-05-21 出版日期:2021-10-15 发布日期:2021-10-18
  • 通讯作者: 李天瑞(trli@swjtu.edu.cn)
  • 作者简介:shihao_zura@163.com
  • 基金资助:
    四川省重点研发项目(2020YFG0035)

Medical Entity Relation Extraction Based on Deep Neural Network and Self-attention Mechanism

ZHANG Shi-hao, DU Sheng-dong, JIA Zhen, LI Tian-rui   

  1. School of Computing and Artificial Intelligence,Southwest Jiaotong University,Chengdu 611756,China
  • Received:2021-03-29 Revised:2021-05-21 Online:2021-10-15 Published:2021-10-18
  • About author:ZHANG Shi-hao,born in 1996,postgraduate.His main research interests include information extraction and natural language processing.
    LI Tian-rui,born in 1969,Ph.D,professor,Ph.D supervisor,is a distinguished member of China Computer Federation.His main research interests include big data intelligence,rough sets and granular computing.
  • Supported by:
    Sichuan Key R&D Project(2020YFG0035).

摘要: 随着医学信息化的推进,医学领域已经积累了海量的非结构化文本数据,如何从这些医学文本中挖掘出有价值的信息,是医学行业和自然语言处理领域的研究热点。随着深度学习的发展,深度神经网络被逐步应用到关系抽取任务中,其中“recurrent+CNN”网络框架成为了医学实体关系抽取任务中的主流模型。但由于医学文本存在实体分布密度较高、实体之间的关系交错互联等问题,使得 “recurrent+CNN”网络框架无法深入挖掘医学文本语句的语义特征。基于此,在“recurrent+CNN”网络框架基础之上,提出一种融合多通道自注意力机制的中文医学实体关系抽取模型,包括:1)利用BLSTM捕获文本句子的上下文信息;2)利用多通道自注意力机制深入挖掘句子的全局语义特征;3)利用CNN捕获句子的局部短语特征。通过在中文医学文本数据集上进行实验,验证了该模型的有效性,其精确率、召回率和F1值与主流的模型相比均有提高。

关键词: 医学文本, 实体关系抽取, 多通道自注意力, 深度学习

Abstract: With the advancement of medical informatization,a large amount of unstructured text data has been accumulated in the medical field.How to mine valuable information from these medical texts is a research hotspot in the field of medical profession and natural language processing.With the development of deep learning,deep neural network is gradually applied to relation extraction task,and “recurrent+CNN” network framework has become the mainstream model in medical entity relation extraction task.However,due to the problems of high entity density and the cross-connection of relationships between entities in medical texts,the “recurrent+CNN” network framework cannot deeply mine the semantic features of medical texts.Based on the “recurrent+CNN” network framework,this paper proposes a Chinese medical entity relation extraction model with multi-channel self-attention mechanism.It includes that BLSTM is used to capture the context information of text sentences,a multi-channel self-attention mechanism is used to mine the global semantic features of sentences,and CNN is used to capture the local phrase features of sentences.The effectiveness of the model is verified by experiments on Chinese medical text dataset.The precision,recall and F1 value of the model are improved compared with the mainstream models.

Key words: Medical text, Entity relation extraction, Multi-channel self-attention, Deep learning

中图分类号: 

  • TP391
[1]GOLSHAN P N,DASHTI H R,AZIZI S,et al.A Study of Recent Contributions on Information Extraction[J].arXiv:1803.05667,2018.
[2]LIU Q,LI Y,DUAN H,et al.Knowledge Graph Construction Techniques[J].Journal of Computer Research and Development,2016,53:582-600.
[3]GRISHMAN R,SUNDHEIM B.Message understanding confe-rence-6:a brief history[C]//Proceedings of the 16th Conference on Computational Linguistics.New York:ACM Press,1996:466-471.
[4]UZUNER O,SOUTH B,SHEN S Y,et al.2010 i2b2/VA challenge on concepts,assertions,and relations in clinical text[J].Journal of the American Medical Informatics Association,2011,18(5):552-556.
[5]NING S M,TENG F,LI T R.Multi-Channel Self-AttentionMechanism for Relation Extraction in Clinical Records[J].Chinese Journal of Computers,2020,43(5):916-929.
[6]HAN X,GAO T Y,LIN Y K,et al.More data,more relations,more context and more openness:a Review and outlook for relation extraction[J].arXiv:2004.03186,2020.
[7]ZHAO S,GRISHMAN R.Extraction relations with integrated information using kernel methods[C]//Proceedings of the 41st Annual Meeting of the Association for Computational Linguistics.Stroudsburg:ACL,2005:419-426.
[8]GUO X Y,HE T T,HU X H,et al.Chinese Named Entity Relation Extraction Based Syntactic and Semantic Features[J].Journal of Chinese Information Processing,2014,28(6):183-189.
[9]KAMBHATLA N.Combining lexical,syntactic,and semanticfeatures with maximum entropy models for extracting relation[C]//Proceedings of ACL on Interactive Poster and Demonstration Sessions.Stroudsburg:ACL,2004:22-26.
[10]ZHOU J.Chinese entity relation extraction based on conditional random fields model[J].Computer Engineering,2010,36(24):192-194.
[11]RINK B,HARABAGIU S,ROBERTS K.Automatic extraction of relations between medical concepts in clinical texts[J].Journal of the American Medical Informatics Association,2011,18(5):594-600.
[12]D'SOUZA J,NG V.Ensemble-Based Medical Relation Classification[C]//25th International Conference on Computational Linguistics.Dublin:COLING,2014:1682-1693.
[13]KIM S,LIU H,YEGANOVA L,et al.Extracting drug-drug interactions from literature using a rich feature-based linear kernel approach[J].Journal of Biomedical Informatics,2015,55(2):23-30.
[14]ZENG D J,LIU K,LAI S W,et al.Relation classification via convolutional deep neural network[C]//Proceedings of the 25th International Conference on Computational Linguistics.Stroudsburg:ACL,2014:2335-2344.
[15]ZHANG D X,WANG D.Relation classification via recurrent neural network[J].arXiv:1508.01006,2015.
[16]ZHANG S,ZHENG D Q,HU X C,et al.Bidirectional Long short-term memory networks for relation classification[C]//Proceedings of the 29th Pacific Asia Conference on Language,Information and Computation.Stroudsburg:ACL,2015:73-78.
[17]ZHU J Z,QIAO J Z,DAI X X,et al.Relation classification via target-concentrated attention CNNs[C]//International Confe-rence on Neural Information Processing.Berlin:Springer,2017:137-146.
[18]ZHOU P,SHI W,TIAN J,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.Stroudsburg:ACL,2016:207-212.
[19]LEE J,SEO S,CHOI Y S.Semantic relation classification via bidirectional LSTM networks with entity-aware attention using latent entity typing[J].arXiv:1901.08163,2019.
[20]CAI R,ZHANG X D,WANG H F.Bidirectional recurrent con-volutional neural network for relation classification[C]//Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics.Stroudsburg:ACL,2016:756-765.
[21]ZHANG X B,CHEN F C,HUANG R Y.A combination of RNN and CNN for attention-based relation classification[J].Procedia Computer Science,2018,131:911-917.
[22]TRAN V H,PHI V T,SHINDO H,et al.Relation Classification Using Segment-Level Attention-based CNN and Dependency-based RNN[C]//Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics.Stroudsburg:ACL,2019:2793-2798.
[23]SAHU S,ANAND A,ORUGANTY K,et al.Relation extraction from clinical texts using domain invariant convolutional neural network[C]//Proceedings of the 15th Workshop on Biomedical Natural Language Processing.2016:206-215.
[24]ZHOU H W,LANG C K,LIU Z,et al.Knowledge-guided con-volutional networks for chemical-disease relation extraction[J].BMC Bioinformatics,2019,20(1):260-273.
[25]SAHU S,ANAND A.Drug-Drug Interaction Extraction from Biomedical Texts Using Long Short-Term Memory Network[J].Journal of Biomedical Informatics,2018,86:15-24.
[26]BAI T,WANG C,WANG Y,et al.A novel deep learning me-thod for extracting unspecific biomedical relation[J].Concurrency and Computation:Practice and Experience,2020,32:1-11.
[27]RAJ D,SAHU S,ANAND A.Learning local and global contexts using a convolutional recurrent network model for relation classification in biomedical text[C]//Proceedings of the 21st Conference on Computational Natural Language Learning.Vancouver:CoNLL,2017:311-321.
[28]HE B,GUAN Y,DAI R.Convolutional Gated Recurrent Units for Medical Relation Classification[C]//2018 IEEE InternationalConference on Bioinformatics and Biomedicine.New York:IEEE Press,2019:646-650.
[29]LIN Z,FENG M,SANTOS C N,et al.A structured self-attentive sentence embedding[J].arXiv:1703.03130,2017.
[1] 董晓梅, 王蕊, 邹欣开. 面向推荐应用的差分隐私方案综述[J]. 计算机科学, 2021, 48(9): 21-35.
[2] 周新民, 胡宜桂, 刘文洁, 孙荣俊. 基于多模态多层级数据融合方法的城市功能识别研究[J]. 计算机科学, 2021, 48(9): 50-58.
[3] 钱梦薇, 过弋. 融合偏置深度学习的距离分解Top-N推荐算法[J]. 计算机科学, 2021, 48(9): 103-109.
[4] 徐涛, 田崇阳, 刘才华. 基于深度学习的人群异常行为检测综述[J]. 计算机科学, 2021, 48(9): 125-134.
[5] 张新峰, 宋博. 一种基于改进三元组损失和特征融合的行人重识别方法[J]. 计算机科学, 2021, 48(9): 146-152.
[6] 林椹尠, 张梦凯, 吴成茂, 郑兴宁. 利用生成对抗网络的人脸图像分步补全法[J]. 计算机科学, 2021, 48(9): 174-180.
[7] 黄晓生, 徐静. 基于PCANet的非下采样剪切波域多聚焦图像融合[J]. 计算机科学, 2021, 48(9): 181-186.
[8] 田野, 陈宏巍, 王法胜, 陈兴文. 室内移动机器人的SLAM算法综述[J]. 计算机科学, 2021, 48(9): 223-234.
[9] 谢良旭, 李峰, 谢建平, 许晓军. 基于融合神经网络模型的药物分子性质预测[J]. 计算机科学, 2021, 48(9): 251-256.
[10] 冯霞, 胡志毅, 刘才华. 跨模态检索研究进展综述[J]. 计算机科学, 2021, 48(8): 13-23.
[11] 王立梅, 朱旭光, 汪德嘉, 张勇, 邢春晓. 基于深度学习的民事案件判决结果分类方法研究[J]. 计算机科学, 2021, 48(8): 80-85.
[12] 郭琳, 李晨, 陈晨, 赵睿, 范仕霖, 徐星雨. 基于通道注意递归残差网络的图像超分辨率重建[J]. 计算机科学, 2021, 48(8): 139-144.
[13] 刘帅, 芮挺, 胡育成, 杨成松, 王东. 基于深度学习SuperGlue算法的单目视觉里程计[J]. 计算机科学, 2021, 48(8): 157-161.
[14] 王施云, 杨帆. 基于U-Net特征融合优化策略的遥感影像语义分割方法[J]. 计算机科学, 2021, 48(8): 162-168.
[15] 田嵩旺, 蔺素珍, 杨博. 基于多判别器的多波段图像自监督融合方法[J]. 计算机科学, 2021, 48(8): 185-190.
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
[1] 夏庆勋,庄毅. 一种基于局部性原理的远程验证机制[J]. 计算机科学, 2018, 45(4): 148 -151 .
[2] 魏芹双,武优西,刘靖宇,朱怀忠. 基于密度约束和间隙约束的对比模式挖掘[J]. 计算机科学, 2018, 45(4): 252 -256 .
[3] 童泽平,李涛,李立杰,任亮. 基于随机需求与产能限制的供应链协同优化研究[J]. 计算机科学, 2018, 45(4): 260 -265 .
[4] 瞿中,赵从梅. 一种抗遮挡的自适应尺度目标跟踪算法[J]. 计算机科学, 2018, 45(4): 296 -300 .
[5] 胡庆成, 张勇, 邢春晓. 基于有重叠社区划分的社会网络影响最大化方法研究[J]. 计算机科学, 2018, 45(6): 32 -35 .
[6] 朱文强. 面向O2O服务的移动社交网络个性化可信群体识别模型[J]. 计算机科学, 2018, 45(6): 76 -83 .
[7] 池凯凯, 林一民, 李燕君, 程珍. 能量捕获传感网中吞吐量最大化的占空比方案[J]. 计算机科学, 2018, 45(6): 100 -104 .
[8] 殷秋实, 陈建华. 多服务器环境下基于椭圆曲线密码的改进的身份认证协议[J]. 计算机科学, 2018, 45(6): 111 -116 .
[9] 刘景玮, 刘京菊, 陆余良, 杨斌, 朱凯龙. 基于网络攻防博弈模型的最优防御策略选取方法[J]. 计算机科学, 2018, 45(6): 117 -123 .
[10] 赖文星, 邓忠民. 基于支配强度的NSGA2改进算法[J]. 计算机科学, 2018, 45(6): 187 -192 .