计算机应用研究2025，Vol.42Issue(11)：3225-3235,11.DOI:10.19734/j.issn.1001-3695.2025.03.0094
基于知识图谱的药物推荐方法研究综述
Review of medication recommendation methods based on knowledge graph
摘要
Abstract
Medication recommendation utilizes an analysis of individual health status,medical history,genetic information,and lifestyle factors to provide personalized medication treatment plans for patients.However,this technology still faces challenges such as data sparsity,cold start problems,and interpretability in practical applications.The knowledge graph,with its rich struc-tured semantic knowledge,serves as auxiliary information for recommendation systems,effectively addressing these issues and enhancing system performance.Therefore,this article reviewed the current development status of medication recommendation methods based on knowledge graphs and their applications in various problems.Firstly,the article systematically outlined the relevant background knowledge,and identified the common and domain-specific issues in medication recommendation.It dis-cussed the advantages and limitations of medication recommendation methods based on knowledge graphs from both problem and technological perspectives,including traditional knowledge graph recommendation methods,recommendation methods for fusing multi-modal knowledge graphs,and recommendation methods for knowledge graphs integrated with large language model.Finally,the article proposed insights into the future development prospects of this field.关键词
知识图谱/推荐系统/药物推荐/多模态知识图谱/大语言模型Key words
knowledge graph(KG)/recommendation system/medication recommendation/multi-modal knowledge graph/large language model分类
计算机与自动化引用本文复制引用
彭琳,汪宇,叶青,程春雷,贺佳..基于知识图谱的药物推荐方法研究综述[J].计算机应用研究,2025,42(11):3225-3235,11.基金项目
国家自然科学基金资助项目(82260988) （82260988）
江西省自然科学基金资助项目(20224BAB206102) （20224BAB206102）
江西省教育厅科学技术研究项目(GJJ2200923) （GJJ2200923）
江西省卫生和计划生育委员会科技计划资助项目(202211404) （202211404）
江西省高校人文社会科学研究项目(JC19125) （JC19125）
江西中医药大学2023年度重点学科建设经费资助课题(2023jzzdxk027) （2023jzzdxk027）
