A Knowledge-enhanced Two-stage Generative Framework for Medical Dialogue Information Extraction

作     者：Zefa Hu Ziyi Ni Jing Shi Shuang Xu Bo Xu Zefa Hu;Ziyi Ni;Jing Shi;Shuang Xu;Bo Xu

作者机构：School of Artificial IntelligenceUniversity of Chinese Academy of SciencesBeijing 100049China Institute of AutomationChinese Academy of SciencesBeijing 100190China 

出 版 物：《Machine Intelligence Research》 (机器智能研究（英文版）)

年 卷 期：2024年第21卷第1期

页      面：153-168页

核心收录：

学科分类：081203[工学-计算机应用技术] 08[工学] 0835[工学-软件工程] 10[医学] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

基　　金：This work was supported by the Key Research Program of the Chinese Academy of Sciences(No.ZDBSSSW-JSC006) the National Natural Science Foundation of China(No.62206294) 

主　　题：Medical dialogue understanding information extraction text generation knowledge-enhanced prompt low-resource setting dataaugmentation 

摘      要：This paper focuses on term-status pair extraction from medical dialogues(MD-TSPE),which is essential in diagnosis dia-logue systems and the automatic scribe of electronic medical records(EMRs).In the past few years,works on MD-TSPE have attracted increasing research attention,especially after the remarkable progress made by generative ***,these generative methods output a whole sequence consisting of term-status pairs in one stage and ignore integrating prior knowledge,which demands a deeper un-derstanding to model the relationship between terms and infer the status of each *** paper presents a knowledge-enhanced two-stage generative framework(KTGF)to address the above *** task-specific prompts,we employ a single model to com-plete the MD-TSPE through two phases in a unified generative form:We generate all terms the first and then generate the status of each generated *** this way,the relationship between terms can be learned more effectively from the sequence containing only terms in the first phase,and our designed knowledge-enhanced prompt in the second phase can leverage the category and status candidates of the generated term for status ***,our proposed special statusnot mentionedmakes more terms available and en-riches the training data in the second phase,which is critical in the low-resource *** experiments on the Chunyu and CMDD datasets show that the proposed method achieves superior results compared to the state-of-the-art models in the full training and low-re-sourcesettings.