首页|图模驱动的在线医疗健康智慧问答服务研究

图模驱动的在线医疗健康智慧问答服务研究

扫码查看
[目的/意义]学者们重视追求医疗智慧问答相关技术本身的前沿性,对基础理论的探讨研究较少,两者未能融合发展.[方法/过程]在辨析相关概念的基础上,首先阐述在线医疗健康领域智慧问答服务的内涵及特征,之后剖析知识图谱与大语言模型的联系及两者的互补融合思路,最后提出图模驱动的在线医疗健康智慧问答服务.[结果/结论]文章将医疗智慧问答服务理论特征贯穿智慧问答服务的全过程,创新性地提出其智慧问答服务应包含大语言模型驱动的医疗知识图谱构建、知识图谱增强的医疗大模型训练、图模驱动的智慧问答服务流程三部分.本研究实现了理论与技术的有机结合,研究成果可用于后续医疗智慧问答的实践性工作.
Research on the Framework of Online Medical Health Wisdom Q&A Services Driven by Knowledge Graph and Large Language Model
[Purpose/Significance]Scholars mostly focused on pursuing the cutting-edge technology of medical intel-ligence Q&A,with less exploration and research on basic theories,and the two have not been integrated and developed.[Method/Process]On the basis of distinguishing relevant concepts,this study first elaborated on the connotation and char-acteristics of wisdom Q&A services in the field of online healthcare.Then,it deeply analyzed the connection between knowledge graphs and large language models,as well as the complementary integration ideas of the two in the field of online healthcare wisdom Q&A.Finally,it proposed a framework of online medical health wisdom Q&A services driven by knowl-edge graphs and large language models.[Result/Conclusion]The article integrates the theoretical characteristics of medi-cal wisdom Q&A services throughout the entire process of wisdom Q&A services,and innovatively proposes that its wisdom Q&A services should include three parts:the construction of medical knowledge graphs driven by large language models,the training of medical large models enhanced by knowledge graphs,and the process of wisdom Q&A services driven by knowledge graphs and large language models.This study achieved a deep integration of theory and technology,and the re-search results can be used for practical work in subsequent medical wisdom Q&A services.

knowledge graphlarge language modelonline medical healthwisdom Q&A services

张君冬、刘江峰、邓景鹏、刘艳华、黄奇

展开 >

南京大学信息管理学院,江苏南京 210023

南京大学数据智能与交叉创新实验室,江苏 南京 210023

武汉大学信息管理学院,湖北武汉 430064

南京中医药大学卫生经济管理学院,江苏南京 210023

展开 >

知识图谱 大语言模型 在线医疗健康 智慧问答服务

2025

现代情报
中国科学技术情报学会 吉林省科技信息研究所

现代情报

北大核心
影响因子:1.133
ISSN:1008-0821
年,卷(期):2025.45(1)