摘要
随着信息技术在中医药领域广泛应用,中医药领域的数据越来越丰富,基于数据驱动的中医药研究应运而生,利用知识图谱对中医药数据进行管理、分析和展示成为中医药传承创新的重要手段。因此,基于中医药知识图谱的推理成为当今中医药研究领域的热点。为更好利用知识图谱推理技术推动中医药的创新和发展,文章在概述中医药知识图谱推理的基础上,重点从基于逻辑规则、分布式表示和神经网络3个方面介绍知识图谱推理的发展及其在中医药领域的应用。最后,文章还总结知识推理方法的优劣,并从知识图谱推理的可解释性研究、时态知识图谱和多模态信息融合3个角度对中医药知识图谱推理的研究与应用进行了展望。
With the widespread application of information technology in the field of Traditional Chinese Medicine(TCM),the data in the TCM domain is becoming increasingly rich.The emergence of data-driven research in TCM has utilized knowledge graphs for the management,analysis,and presentation of TCM data,becoming a crucial method for the inheritance and innovation of Traditional Chinese Medicine.Therefore,reasoning based on TCM knowledge graphs has become a hot topic in contemporary TCM research.In order to better utilize knowledge graph reasoning technology to promote innovation and development in TCM,this article,after providing an overview of TCM knowledge graph reasoning,focuses on the development of knowledge graph reasoning in TCM and its applications from three perspectives:logic rule-based,distributed representation,and neural network-based methods.Finally,the article summarizes the advantages and disadvantages of knowledge reasoning methods and provides prospects for research and applications in TCM knowledge graph reasoning,including studies on interpretability,temporal knowledge graphs,and the fusion of multimodal information.
作者
刘海煜
瞿小龙
沈翊康
蒋卓斌
谢晓霞
赵玉凤
张小平
LIU Haiyu;QU Xiaolong;SHEN Yikang;JIANG Zhuobin;XIE Xiaoxia;ZHAO Yufeng;ZHANG Xiaoping(National Data Center of Tradition Chinese Medicine,China Academy of Chinese Medical Sciences,Beijing 100700,China;Beijing Forestry University,Beijing 100083,China)
出处
《世界科学技术-中医药现代化》
北大核心
2025年第3期601-611,共11页
Modernization of Traditional Chinese Medicine and Materia Medica-World Science and Technology
基金
国家自然科学基金委员会面上项目(82374621):基于多模态知识图谱推理的胃癌前疾病风险预警方法研究,负责人:张小平。
关键词
中医药
知识图谱推理
逻辑规则
分布式表示
神经网络
多模态
Traditional Chinese medicine
Knowledge graph reasoning
Logic rules
Distributed representation
Neural network
Multimodal
