LIU Ziqi,HU Jiancheng,MOU Gufang.Classification of Screening Criteria for Chinese Clinical Trials based on Deep Learning[J].Journal of Chengdu University of Information Technology,2024,39(02):170-177.[doi:10.16836/j.cnki.jcuit.2024.02.007]
基于深度学习的中文临床实验筛选标准的分类
- Title:
- Classification of Screening Criteria for Chinese Clinical Trials based on Deep Learning
- 文章编号:
- 2096-1618(2024)02-0170-08
- 分类号:
- TP391.1
- 文献标志码:
- A
- 摘要:
- 针对大多数临床实验筛选标准的分类研究都集中在英文资格标准上,研究适合中文资格标准的分类模型,利用第五届中国健康信息处理会议开发的中文临床实验短文本数据集,结合神经网络和预训练语言模型对分类任务进行构建和微调,比较分析Word2vec-BiLSTM模型、CNN模型、RNN模型、预训练语言模型在此应用上的效果差异,并通过实验得到预训练模型ERNIE的分类效果优于其他模型。针对数据不平衡这一特征,对数量较少的类别语料进行数据增强后可有效提升模型的性能和效果,结果显示ERNIE模型的宏观平均F1值和微观平均F1值分别可达到0.8281和0.8537。
- Abstract:
- Classification research for most clinical trial screening criteria focuses on English eligibility criteria. This paper compares the characteristic of classification models suitable for Chinese eligibility criteria, using the Chinese clinical trial short text dataset developed by the 5th China Health Information Processing Conference, combined with neural networks and pre-trained language models to construct classification tasks and fine-tuning, analyzed the differences in the effects of the Word2vec-BiLSTM model, CNN model, RNN model, and pre-trained language model in this application, and obtained through experiments that the classification effect of the pre-trained model ERNIE performsbetter. In view of the characteristic of data imbalance, the performance and effect of the model can be effectively improved after data enhancement of a small number of category corpora. The results show that the macro-average F1 value and micro-average F1 value of the ERNIE model can reach 0.8281 and 0.8537, respectively.
参考文献/References:
[1] 左圆圆,王媛媛,蒋珊珊, 等.数据可视化分析综述[J].微计算信息,2019,(1):82-83.
[2] 王勇,李帅.自然语言处理在医学文本挖掘中的应用[J].电子技术与软件工程,2019,(7):190-190.
[3] 宗辉,张泽宇,杨金璇,等.基于人工智能的中文临床试验筛选标准文本分类研究[J].生物医学工程杂志,2021,38(1):105-121.
[4] Hao T,Rusanov A,Boland MR,et al.Clustering clinical trials with similar eligibility criteria features[J].Journal of biomedical informatics,2014,52(C):112-120.
[5] Regev Y,Finkelstein M,Feldman R,et al.Rule-based extraction of experimental evidence in the biomedical domain:the KDD Cup 2002(task 1)[J].SIGKDD Explor.Newsl,2002,4(2):90-92.
[6] Luo Z,Stephen B,Weng C,et al.Semi-automatically inducing semantic classes of clinical research eligibility criteria using UMLS and hierarchical clustering[C].AMIA Annual Symposium Proceedings:American Medical Informatics Association,2010,2010:487.
[7] Luo Z, Yetisgen-Yildiz M, Weng C. Dynamic categorization of clinical research eligibility criteria by hierarchical clustering[J].Journal of biomedical informatics,2011,44(6):927-935.
[8] Marafino B J,Davies J M,Bardach N S,et al.N-gram support vector machines for scalable procedure and diagnosis classification,with applications to clinical free text data from the intensive care unit[J].Journal of the American Medical Informatics Association,2014,21(5): 871-875.
[9] Yi K,Beheshti J.A hidden Markov model-based text classification of medical documents[J].Journal of Information Science,2009,35(1):67-81.
[10] 王培,王亚文,卢苗苗.基于BERT模型的中医文本分类研究[J].电脑知识与技术,2021,17(27):13-20.
[11] 李启行,廖薇.基于注意力机制的生物医学文本分类模型[J].中国医学物理学杂志,2022,39(4):518-523.
[12] 钟桂凤,庞雄文,隋栋.基于Word2Vec和改进注意力机制AlexNet-2的文本分类方法[J].计算机科学, 2022, 49(4):288-293.
[13] Kim Y.Convolutional Neural Networks for Sentence Classification[J/OL].https://ui.adsabs.harvard.edu/abs/2014arXiv1408.5882K,2014, 08, 01/2022,12,11,2022.
[14] Tan M,Santos C,Xiang B,et al.Lstm-based deep learning models for non-factoid answer selection[J/OL].https://ui.adsabs.harvard.edu/abs/2015arXiv151104108T,2015,11,01/2022,12,11,2022.
[15] Zhou P,Shi W,Tian J,et al.Attention-based bidirectional long short-term memorynetworks for relation classification[C].Proceedings of the 54th annual meeting of the association for computational linguistics(volume 2:Short papers).2016:207-212.
[16] Cai L,Zhou S,Yan X,et al.A stacked BiLSTM neural network based on coattention mechanism for question answering[J].Computational intelligence and neuroscience,2019,2019:9543490.
[17] Vaswani A,Shazeer N,Parmar N,et al.Attention is all you need[J/OL].https://ui.adsabs.harvard.edu/abs/2017 arXiv170603762V,2017,01,01/2022,12,11,2022.
[18] Devlin J,Chang M W,Lee K,et al.Bert:Pre-training of deep bidirectional transformers for language understanding[J/OL].https://ui.adsabs.harvard.edu/abs/2018arXiv181004805D,2018,10,01/2022,12,11,2022.
[19] Sun Y,Wang S,Li Y,et al.ERNIE:Enhanced Representation through Knowledge Integration[J/OL].https://ui.adsabs.harvard.edu/abs/2019arXiv190409223S,2019,05,01/2022,12,11,2022.
[20] 第五届中国健康信息处理会议(CHIP 2019).评测三:临床试 验筛选标准短文本分类[EB/OL].http://www.cips-chip.org.cn:8088/evaluation,2020,11,06/2022,12,11,2022.
[21] Wei J,Zou K.Eda:Easy data augmentation techniques for boosting performance on text classification tasks[J/OL].https://ui.adsabs.harvard.edu/abs/2019arXiv190111196W,2019,01,01/2022,12,11,2022.
备注/Memo
收稿日期:2022-12-28