计算机工程与应用 ›› 2023, Vol. 59 ›› Issue (19): 40-51.DOI: 10.3778/j.issn.1002-8331.2301-0006

• 热点与综述 • 上一篇    下一篇

融合医疗知识图谱的推荐系统研究进展

沈希宇,蔡肖红,曹慧   

  1. 山东中医药大学 智能与信息工程学院,济南 250355
  • 出版日期:2023-10-01 发布日期:2023-10-01

Research Progress of Recommendation System Based on Medical Knowledge Graph

SHEN Xiyu, CAI Xiaohong, CAO Hui   

  1. College of Intelligence and Information Engineering, Shandong University of Traditional Chinese Medicine, Jinan 250355, China
  • Online:2023-10-01 Published:2023-10-01

摘要: 医疗知识图谱因其结构化的语义知识特点,可以为推荐系统提供新型的辅助信息。推荐系统与医疗知识图谱相结合,不仅能有效缓解数据稀疏等问题,还增强了推荐结果的准确性以及可解释性,从而实现医疗信息的个性化推荐。针对医疗领域专业壁垒坚固、概念术语繁多等特点,对医疗知识图谱架构进行了系统梳理;总结传统推荐算法,并对比分析了其优缺点;对基于路径、基于嵌入和基于融合的推荐系统分别总结介绍,重点对结合医疗实践的研究成果及其优缺点进行归纳总结;对具可行性的未来研究方向进行了展望。

关键词: 知识图谱, 智能医疗, 推荐系统, 自然语言处理

Abstract: Medical knowledge graph can provide new auxiliary information for recommendation system because of its structural semantic knowledge characteristics. The combination of recommendation system and medical knowledge graph can not only effectively alleviate the problem of sparse data, but also enhance the accuracy and interpretability of recommendation results, thus realizing personalized recommendation of medical information. Aiming at the characteristics of strong professional barriers and various conceptual terms in the medical field, the medical knowledge graph structure is systematically sorted out. This paper summarizes the traditional recommendation algorithms and compares their advantages and disadvantages. It summarizes and introduces the recommendation systems based on path, embedding and fusion respectively, focusing on the research results of combining medical practice and their advantages and disadvantages. The existing problems and challenges in current research are analyzed, and the feasible future research direction is prospected.

Key words: knowledge graph, intelligent medicine, recommendation system, natural language processing