Based on the BERT-BiLSTM-CRF, this paper proposed a knowledge extraction model that extracts relations firstly and entities secondly; the output of the relation recognition model is the input of the entity recognition model. The experimental conclusions show that the knowledge extraction model can effectively extract relations and entities of Dictionary of Chinese Ethnic Medicine. The knowledge graph of Chinese ethnic medicine extracted by the model is constructed, and the application of knowledge visualization and intelligent question answering are realized.
随着谷歌知识图谱[4]的发展,知识图谱在医疗领域里应用不断向前推进,如中医养生知识图谱[5]、医药知识图谱[6]、中医药知识图谱[7]、生物医药知识图谱[8]、心脏病中文知识图谱[9]、中文症状知识库[10];在其知识智能应用方面,有面向患者的智能医生问答[11]、辅助医生诊断的平行胃肠诊疗系统[12]、融合知识图谱与深度学习的药物发现[13]、针对特定疾病和特定医疗领域的人工智能辅助诊疗专家诊断系统(如IBM的Waston[14]),以及近期基于BERT(bidirectional encoder representation from transformers)融合大规模疾病术语和常用语[15]的应用研究。
本研究完成的工作主要有: ① 基于BERT-BiLSTM-CRF构建了知识抽取模型,应用到《中国民族药辞典》的关系和实体识别中,进行知识抽取,第一个关系识别模型的输出是第二个实体识别模型的输入,这两个模型共同组合成知识抽取模型,其在测试集上取得了很好的性能;② 利用从《中国民族药辞典》中抽取的知识,设计了民族药知识图谱模式,构建了民族药知识图谱,基于民族药知识图谱探索了民族药相关智能应用,实现了民族药知识图谱可视化和智能问答。
HUANGF K. Developmental status of Chinese ethnomedicine and integrative stratagem [J]. Journal of South-Central University for Nationalities (Natural Science Edition), 2008, 27(2): 36-39. DOI: 10.3969/j.issn.1672-4321.2008.02.011(Ch ).
ZHUG B. Selection between Chinese traditional medicine and western alternative medicine [J]. World Science and Technology-Modernization of Traditional Chinese Medicine, 2001, 3(5): 58-62. DOI:10.3969/j.issn.1674-3849.2001.05.015(Ch ).
ZHUF, SHAOR, WUY P. Industry development status and strategies of Chinese Ethnic Medicine [J]. Modern Business Trade Industry, 2013, 25(10): 4-5. DOI:10.3969/j.issn.1672-3198.2013.10.002(Ch ).
[7]
AMITS. Introducing the Knowledge Graph [EB/OL]. [2012-05-16]. https://
YUT, LIJ H, YUQ, et al. The construction and application of knowledge mapping of health preservation of traditional Chinese medicine [J]. China Digital Medicine, 2017, 12(12): 64-66. DOI:10.3969/j.issn.1673-7571.2017.12.020(Ch ).
HUANGW L. Research on the Construction of Medical Knowledge Graph QA System Based on Deep Learning [D]. Wuhan:Huazhong University of Science & Technology,2019 (Ch). DOI: 10.21661/r-497961 .
RUANT, SUNC L, WANGH F, et al. Construction of traditional Chinese medicine knowledge graph and its application [J]. Journal of Medical Intelligence, 2016, 37(4): 8-13. DOI:10.3969/j.issn.1673-6036.2016.04.002(Ch ).
FUY, LIUM F, QIAOR. Construction of Chinese knowledge graph of heart disease [J].Journal of Wuhan University(Natural Science Edition),2020,66(3):261-267. DOI:10.14188/j.1671-8836.2018.0217(Ch ).
ZANH Y, HANY C, FANY X, et al. Construction and analysis of symptom knowledge base in Chinese [J]. Journal of Chinese Information Processing, 2020,34(4):30-37. DOI: 10.3969/j.issn.1003-0077.2020.04.004(Ch ).
XIEG, WUG W, RENJ H, et al. Research on intelligent doctor framework for patient [J].Journal of Frontiers of Computer Science and Technology,2018,12(9):1475-1486. DOI: 10.3778/j.issn.1673-9418.1806024(Ch ).
SANGS T, YANGZ H, LIUX X, et al. A method combining knowledge graph and deep learning for drug discovery [J].Pattern Recognition and Artificial Intelligence,2018,31(12):1103-1110. DOI :10.16451 /j.cnki.issn1003-6059.201812005(Ch ).
[26]
STRICKLANDE. IBM Watson, heal thyself: How IBM overpromised and under delivered on AI health care [J]. IEEE Spectrum, 2019, 56(4): 24-31. DOI: 10.1109/MSPEC.2019.8678513 .
ZHANGC T, ZHANGJ Y, ZHANGZ X, et al. Construction of large-scale disease terminology graph with common terms [J].Journal of Computer Research and Development,2020,57(11):2467-2477. DOI:10.7544/issn1000-1239.2020.20190747(Ch ).
JIAM R, ZHANGY, YANZ Y, et al. Species and use of current chinese minority medicine [J].World Science and Technology-Modernization of Traditional Chinese Medicine and Materia Medica, 2015(7): 1546-1550.DOI :10.11842/wst.2015.07.035 (Ch ).
PENGC Y, ZHANGH, BAOL Y, et al. Biological named entity recognition based on conditional random fields [J]. Computer Engineering, 2009, 35(22): 197-199. DOI: 10.3969/j.issn.1000-3428.2009.22.067(Ch ).
[33]
LAMPLEG, BALLESTEROSM, SUBRAMANIANS, et al. Neural architectures for named entity recognition[C]// Proceedings of the 2016 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. San Diego: Association for Computational Linguistics,2016: 260–270. DOI:10.18653/v1/N16-1030 .
[34]
CHIUJ P C, NICHOLSE. Named entity recognition with bidirectional LSTM-CNNs [J]. Transactions of the Association for Computational Linguistics, 2016, 4: 357-370. DOI:10.1162/tacl_a_00104 .
LIL S, GUOY K. Biomedical named entity recognition with CNN-BLSTM-CRF [J]. Journal of Chinese Information Processing, 2018, 32(1): 116-122. DOI: 10.3969/j.issn.1003-0077.2018.01.015(Ch ).
[37]
MINTZM, BILLSS, SNOWR, et al. Distant Supervision for Relation Extraction without Labeled Data [DB/OL] .[2020-10-12].jsessionid=D9E667ECBF5D1A1D5C853026BB9E56F7?doi=10.1.1.148.3346&rep=rep1&type=pdf. DOI:10.3115/1690219.1690287 .
EH H, ZHANGW J, XIAOS Q, et al. Survey of entity relationship extraction based on deep learning [J]. Journal of Software,2019,30(6):1793-1818. DOI: 10.13328/j.cnki.jos.005817(Ch ).
[42]
WANGQ, WANGT, XUC. Using a knowledge graph for hypernymy detection between Chinese symptoms[C]// 2018 Tenth International Conference on Advanced Computational Intelligence (ICACI). New York: IEEE Press, 2018: 300-307. DOI: 10.1109/icaci.2018.8377528 .
LIUS W, SHAOY F, QIANL H. Biomedical causality relation extraction based on joint learning [J]. Journal of Chinese Information Processing,2020,34(4):60-68. DOI: 10.3969/j.issn.1003-0077.2020.04.008(Ch ).
[45]
DEVLINJ, CHANGM W, LEEK, et al. Bert: Pre-training of deep bidirectional transformers for language understanding [C]// Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers). Minneapolis: Association for Computational Linguistics,2019:4171-4186. DOI: 10.18653/v1/N19-1423 .
[46]
SOUZAF, NOGUEIRAR, LOTUFOR. Portuguese Named Entity Recognition Using BERT-CRF [DB/OL].[2020-09-12]. https://arxiv.org/pdf/1909.10649.pdf.
[47]
ZHANGW T, JIANGS H, ZHAOS, et al. A BERT-BiLSTM-CRF model for Chinese electronic medical records named entity recognition[C]// 2019 12th International Conference on Intelligent Computation Technology and Automation (ICICTA). Piscataway: IEEE, 2019: 166-169. DOI:10.1109/ICICTA49267.2019.00043 .
[48]
STRAKOVÁJ, STRAKAM, HAJIČJ. Neural architectures for nested NER through linearization [C]// Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics. Florence: Association for Computational Linguistics, 2019: 5326-5331. DOI: 10.18653/v1/p19-1527 .
[49]
CHENC C, CHENK, HSUC Y, et al. Developing guideline-based decision support systems using Protégé and Jess [J]. Computer Methods and Programs in Biomedicine, 2011, 102(3): 288-294. DOI: 10.1016/j.cmpb.2010.05.010 .
[50]
KNUBLAUCHH, FERGERSONR W, NOY N F, et al. The Protégé OWL plugin: An open development environment for semantic web applications [C]//International Semantic Web Conference. Berlin: Springer, 2004: 229-243. DOI:10.1007/978-3-540-30475-3_17 .
[51]
VUKOTICA, WATTN, ABEDRABBOT,et al. Neo4j in Action [M]. Shelter Island :Manning Publications Co, 2014:52-148.