Named Entity Recognition model for Electronic Medical Records Based on External Knowledge Integration and Prompt Learning

Corresponding author: caozijia, ydfyczj@126.com

DOI: 10.12201/bmr.202608.00018

•  

  Abstract: Purpose/Significance To enhance the recognition capability of medical entities in electronic medical records, provide foundational support for clinical decision support and medical knowledge base construction, and facilitate hospital informatization. Method/Process named entity recognition model for electronic medical records based on external knowledge integration and prompt learning is proposed.First, learnable domain-specific prompt vectors initialized from category description texts are constructed and concatenated with the original text to generate domain-focused word embeddings via the ERNIE model. Second, the softLexicon method and an attention mechanism are employed to integrate external dictionary, enhancing the model’s ability to represent ambiguous entities and boundaries. Finally, a two-stage recognition mechanism is adopted, where candidate span pruning and hard negative sample generation improve boundary detection accuracy, and an attention mechanism is used to enable interaction between candidate spans and prompt vectors for classification.Results/Conclusion Comparative experiments and ablation studies were conducted on four publicly available electronic medical record datasets, and the experimental results demonstrate the effectiveness of the proposed model.

  Key words: named entity recognition; prompt learning; external knowledge integration; two-stage recognition mechanism

  Submit time: 7 August 2026

  Copyright: The copyright holder for this preprint is the author/funder, who has granted biomedRxiv a license to display the preprint in perpetuity.
• html

• 图表

• chenjieqing, zhangfeng. Named Entity Recognition in Chinese Electronic Medical Records Using Knowledge Graph Construction. 2023. doi: 10.12201/bmr.202312.00011

  xiaoxiaoxia. Research on named entity recognition of Chinese medical records based on BERT-BiLSTM-CRF with Chinese radicals. 2023. doi: 10.12201/bmr.202303.00004

  xianhongxin, shenghui, masufen, caixiaohong, zhangfengcong, chenyifan, wangpingping. Research Review on Named Entity Recognition of Ancient Chinese Medicine Books Based on Deep Learning. 2026. doi: 10.12201/bmr.202602.00050

  shenrongrong, xiashuaishuai, yanjunfeng. Review on Research of Named Entity Recognition in Chinese Medicine. 2022. doi: 10.12201/bmr.202207.00038

  Deng Jiale, Hu Zhensheng, Lian Wanmin, Hua Yunpeng, Zhou Yi. Research on entity recognition of liver cancer electronic medical records based on RoBERTa-CRF. 2023. doi: 10.12201/bmr.202303.00027

  renhuiling, lixiaoying, wangweijie, wangxu, zhangying. Research on Chinese electronic medical record entity mapping method by fusing similarity algorithm and pre-trained model. 2023. doi: 10.12201/bmr.202305.00015

  wuhuan, hekunlun. Construction of general medical knowledge graph based on evidence-based medicine and electronic medical record data. 2024. doi: 10.12201/bmr.202409.00027

  WU Shengnan, WU Jiahui, DONG Jizong, JIANG Huanyu, WANG Luqi, WANG Xinyao. Research on predicting medical entity relationships based on bipartite network representation learning. 2024. doi: 10.12201/bmr.202407.00041

  Liang Fei, Huang Huai Yu, Li Ren Yuan, Chen Xing. Research on Intelligent Recognition and Intervention Strategies for Quality Control Nodes in the Full Process of Electronic Medical Records from a Data Chain Perspective. 2026. doi: 10.12201/bmr.202610.00045

  gaiyanrong, zhangyunqiu, zhanghui, lichencheng, lujunrui. Research on Quality Analysis and Governance Strategies of Real-World Chinese Electronic Medical Records Data for Knowledge Extraction. 2025. doi: 10.12201/bmr.202511.00077

• ID	Submit time	Number	Download
  1	2026-04-03	10.12201/bmr.202608.00018V1	Download

• Public  Anonymous  To author only Submit Comments