计算机科学 ›› 2024, Vol. 51 ›› Issue (6A): 230500203-7.doi: 10.11896/jsjkx.230500203
尹宝生, 周澎
YIN Baosheng, ZHOU Peng
摘要： 医学领域命名实体识别是信息抽取任务重要的研究内容之一,其训练数据主要来源于临床实验数据、健康档案、电子病历等非结构化文本,然而标注这些数据需要专业人员耗费大量人力、物力和时间资源。在缺乏大规模医学训练数据的情况下,医学领域命名实体识别模型很容易出现识别错误的情况。为解决这一难题,文中提出了一种融合标签知识的中文医学命名实体识别方法,即通过专业领域词典获得文本标签的释义后,分别将文本、标签及标签释义编码,基于自适应融合机制进行融合,有效平衡特征提取模块和语义增强模块的信息流,从而提高模型性能。其核心思想在于医学实体标签是通过总结归纳大量医学数据得到的,而标签释义是对标签进行科学解释和说明的结果,模型融入这些蕴含了丰富的医学领域内的先验知识,可以使其更准确地理解实体在医学领域中的语义并提升其识别效果。实验结果表明,该方法在中文医学实体抽取数据集(CMeEE-V2) 3个基线模型上分别取得了0.71%,0.53%和1.17%的提升,并且为小样本场景下的实体识别提供了一个有效的解决方案。
中图分类号:
| [1]LIN H,LU Y,TANG J,et al.A Rigorous Study on Named Entity Recognition:Can Fine-tuning Pretrained Model Lead to the Promised Land?[C]//Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing(EMNLP).2020:7291-7300. [2]LEE J,YOON W,KIM S,et al.BioBERT:a pre-trained biome-dical language representation model for biome-dical text mining[J].Bioinformatics,2019,36(4):1234-1240. [3]MI F,ZHOU W,CAI F,et al.Self-training improves pre-trai-ning for few-shot learning in task-oriented dialog systems[C]//Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing(EMNLP).2021:1887-1898. [4]HA S,KERSNER M,KIM B,et al.Marionette:Few-shot facereenactment preserving identity of unseen targets[C]//Proceedings of the AAAI Conference on Artificial Intelligence.2020:10893-10900. [5]WANG Y,YAO Q,KWOK J T,et al.Generalizing from a Few Examples:A Survey on Few-shot Learning[J].ACM computing surveys(csur),2020,53(3):1-34. [6]LIUH,TAM D,MUQEETH M,et al.Few-Shot Parameter-Efficient Fine-Tuning is Better and Cheaper than In-Context Learning[J].Advances in Neural Information Processing Systems,2022,35:1950-1965. [7]OSAHOR U,NASRABADI N M.Ortho-shot:low displacement rank regularization with data augmentation for few-shot learning[C]//Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision.2022:2200-2209. [8]SUN Q,LIU Y,CHUA T S,et al.Meta-Transfer Learning for Few-Shot Learning[C]//Proceedings of the IEEE/CVF Confe-rence on Computer Vision and Pattern Recognition(CVPR).2019:403-412. [9]LUO X,XU J,XU Z.Channel importance matters in few-shotimage classification[J].In International Conference on Machine Learning(PMLR),2022,162:14542-14559. [10]DIXIT M,KWITT R,NIETHAMMER M,et al.AGA:Attri-bute-Guided Augmentation[J].Proceedings of the IEEE Confe-rence on Computer Vision and Pattern Recognition,2017,35:7455-7463. [11]SCHWARTZ E,KARLINSKY L,SHTOK J,et al.Delta-en-coder:an effective sample synthesis method for few-shot object recognition[J].Advances in neural information processing systems,2018,31:2850-2860. [12]CHEN J,LIU Q,LIN H,et al.Few-shot named entity recognition with self-describing networks[J].arXiv:2203.12252,2022. [13]LAI P,YE F,ZHANG L,et al.PCBERT:Parent and ChildBERT for Chinese Few-shot NER[C]//Proceedings of the 29th International Conference on Computational Linguistics.2022:2199-2209. [14]MA T,JIANG H,WU Q,et al.Decomposed Meta-Learning for Few-Shot Named Entity Recognition[J].arXiv:2204.05751,2022. [15]WANG J,WANG C,TAN C,et al.SpanProto:A Two-stageSpan-based Prototypical Network for Few-shot Named Entity Recognition[C]//Proceedings of the 2022 Conference on Empi-rical Methods in Natural Language Processing.2022:3466-3476. [16]LI J,CHIU B,FENG S,et al.Few-shot named entity recognition via meta-learning[J].IEEE Transactions on Knowledge and Data Engineering,2020,34(9):4245-4256. [17]MA J,BALLESTEROS M,DOSS S,et al.Label semantics for few shot named entity recognition[J].arXiv:2203.08985,2022. [18]FRITZLER A,LOGACHEVA V,KRETOV M.Few-shot classification in named entity recognition task[C]//Proceedings of the 34th ACM/SIGAPP Symposium on Applied Computing.2019:993-1000. [19]LI J,SUN A,HAN J,LI C.A survey on deep learning for named entity recognition[J].IEEE Transactions on Knowledge and Data Engineering,2020,34(1):50-70. [20]CHURCH K W.Word2Vec[J].Natural Language Engineering,2017,23(1):155-162. [21]PENNINGTON J,SOCHER R,MANNING C D.Glove:Global vectors for word representation[C]//Proceedings of the 2014 Conference on Empirical Methods in Natural Language Proces-sing(EMNLP),2014:1532-1543. [22]DEVLIN J,CHANG M W,LEE K,et al.Bert:Pre-training of deep bidirectional transformers for language understanding[J].Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics(NAACL),2019,21:4171-4186. [23]LU W,LI J,WANG J,et al.A CNN-BiLSTM-AM method forstock price prediction[J].Neural Computing and Applications,2021,33:4741-4753. [24]LAFFERTY J,MCCALLUM A,PEREIRA F.Conditional random fields:probabillstic models for segmenting and labeling sequence data[C]//Proceedings of the 18th International Confe-rence on Machine Learning.2001:282-289. [25]ZAN H Y,LI W X,ZHANG K L,et al.Building a PediatricMedical Corpus:Word Segmentation and Named Entity Annotation[C]//The 21st Chinese Lexical Semantics Workshop.2021:652-664. [26]LU Y,LIU Q,DAI D,et al.Unified structure generation for universal information extraction[C]//Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics(ACL).2022:5755-5772. [27]LI Y,LIU L,SHI S.Empirical analysis of unlabeled entity pro-blem in named entity recognition[J].arXiv:2012.05426,2020. [28]FU Y,LIN N,YANG Z,et al.Towards Malay named entity re-cognition:an open-source dataset and a multi-task framework[J].Connection Science,2023:35(1):2159014. [29]LI X,FENG J,MENG Y,et al.A unified MRC framework fornamed entity recognition[J].arXiv:1910.11476,2019. |
|
||
首页