《计算机应用》唯一官方网站 ›› 2025, Vol. 45 ›› Issue (4): 1157-1168.DOI: 10.11772/j.issn.1001-9081.2024050573

• 人工智能 • 上一篇    下一篇

基于知识图谱和对话结构的问诊推荐方法

徐春, 吉双焱(), 马欢, 孙恩威, 王萌萌, 苏明钰   

  1. 新疆财经大学 信息管理学院,乌鲁木齐 830012
  • 收稿日期:2024-05-09 修回日期:2024-07-09 接受日期:2024-07-11 发布日期:2024-07-25 出版日期:2025-04-10
  • 通讯作者: 吉双焱
  • 作者简介:徐春(1977—),女,贵州毕节人,教授,博士,CCF会员,主要研究方向:自然语言处理、大数据分析
    吉双焱(1999—),女,山西运城人,硕士研究生,主要研究方向:自然语言处理、医疗数据分析
    马欢(1980—),女(回族),青海西宁人,讲师,硕士,主要研究方向:自然语言处理、医疗数据分析
    孙恩威(1998—),男,安徽马鞍山人,硕士研究生,主要研究方向:自然语言处理、医疗数据分析
    王萌萌(1998—),女,四川南充人,硕士研究生,主要研究方向:自然语言处理、大数据分析
    苏明钰(2000—),女,辽宁鞍山人,硕士研究生,主要研究方向:自然语言处理、大数据分析。
  • 基金资助:
    国家自然科学基金资助项目(62266041);新疆自然科学基金资助项目(2023D01A73)

Consultation recommendation method based on knowledge graph and dialogue structure

Chun XU, Shuangyan JI(), Huan MA, Enwei SUN, Mengmeng WANG, Mingyu SU   

  1. School of Information Management,Xinjiang University of Finance and Economics,Urumqi Xinjiang 830012,China
  • Received:2024-05-09 Revised:2024-07-09 Accepted:2024-07-11 Online:2024-07-25 Published:2025-04-10
  • Contact: Shuangyan JI
  • About author:XU Chun, born in 1977, Ph. D., professor. Her research interests include natural language processing, big data analysis.
    JI Shuangyan, born in 1999, M. S. candidate. Her research interests include natural language processing, medical data analysis.
    MA Huan, born in 1980, M. S., lecturer. Her research interests include natural language processing, medical data analysis.
    SUN Enwei, born in 1998, M. S. candidate. His research interests include natural language processing, medical data analysis.
    WANG Mengmeng, born in 1998, M. S. candidate. Her research interests include natural language processing, big data analysis.
    SU Mingyu, born in 2000, M. S. candidate. Her research interests include natural language processing, big data analysis.
  • Supported by:
    National Natural Science Foundation of China(62266041);Natural Science Foundation of Xinjiang(2023D01A73)

摘要:

针对现有的问诊推荐方法未能充分利用医患间丰富的对话信息和无法捕捉患者实时的健康需求和偏好的问题,提出一种基于知识图谱和对话结构的问诊推荐方法(KGDS)。首先,构建包含评论情感分析和医学专业知识的医疗知识图谱(KG),增强医生和患者的细粒度特征表示;其次,在患者表示学习部分,设计一种患者查询编码器,从词级和句级这2个层面提取查询文本的关键特征,并通过注意力机制加强医患向量间的高阶特征交互;再次,建模诊断对话,充分利用医患间丰富的对话信息增强医患特征表示;最后,设计基于对比学习的对话模拟器,捕捉患者的动态需求和实时偏好,利用模拟的对话表示辅助推荐得分的预测。在真实数据集上的实验结果表明,KGDS相较于最优基线方法在曲线下面积(AUC)、平均值倒数秩(MRR@15)、推荐多样性(Diversity@15)、调和平均值(F1@15)、命中率(HR@15)和归一化折损累计增益(NDCG@15)上分别提高了1.82、1.78、3.85、3.06、10.02和4.51个百分点,验证了KGDS的有效性,且可见情感分析和KG的纳入增强了推荐结果的可解释性。

关键词: 知识图谱, 对话结构, 问诊推荐, 评论情感分析, 注意力机制, 可解释性

Abstract:

Aiming at the problems that the existing consultation recommendation methods do not fully utilize the rich dialogue information between doctors and patients and do not capture patients’ real-time health needs and preferences, a consultation recommendation method based on Knowledge Graph and Dialogue Structure (KGDS) was proposed. Firstly, a medical Knowledge Graph (KG) including comment sentiment analysis and professional medical knowledge was constructed to improve the fine-grained feature representations of doctors and patients. Secondly, in the patient representation learning part, a patient query encoder was designed to extract key features of query text at both word and sentence levels, and to improve the higher-level feature interactions between doctor and patient vectors through attention mechanism. Thirdly, a diagnosis dialogue was modeled to make full use of the rich dialogue information between doctors and patients to enhance the doctor-patient feature representation. Finally, a dialogue simulator based on contrastive learning was designed to capture the dynamic needs and real-time preferences of patients, and the simulated dialogue representation was used to support recommendation score prediction. Experimental results on a real dataset show that compared with the optimal baseline method, KGDS increases AUC(Area Under the Curve), MRR@15(Mean Reciprocal Rank), Diversity@15, F1@15, HR@15 (Hit Ratio) and NDCG@15(Normalized Discounted Cumulative Gain) by 1.82, 1.78, 3.85, 3.06, 10.02 and 4.51 percentage points, respectively, which verifies the effectiveness of the proposed consultation recommendation method, and it can be seen that adding sentiment analysis and KG improves the interpretability of the recommendation results.

Key words: Knowledge Graph (KG), dialogue structure, consultation recommendation, comment sentiment analysis, attention mechanism, interpretability

中图分类号: