Journal of Computer Applications ›› 2024, Vol. 44 ›› Issue (10): 3252-3259.DOI: 10.11772/j.issn.1001-9081.2023101508

• The 40th CCF National Database Conference (NDBC 2023) • Previous Articles     Next Articles

Recommendation method using knowledge graph embedding propagation

Beijing ZHOU1, Hairong WANG1,2(), Yimeng WANG1, Lisi ZHANG1, He MA1   

  1. 1.School of Computer Science and Engineering,North Minzu University,Yinchuan Ningxia 750021,China
    2.The Key Laboratory of Images & Graphics Intelligent Processing of State Ethnic Affairs Commission (North Minzu University),Yinchuan Ningxia 750021,China
  • Received:2023-11-06 Revised:2023-12-23 Accepted:2023-12-28 Online:2024-10-15 Published:2024-10-10
  • Contact: Hairong WANG
  • About author:ZHOU Beijing, born in 1997, M. S. candidate. His research interests include recommendation system based on knowledge graph.
    WANG Yimeng, born in 2000, M. S. candidate. Her research interests include multi-modal recommendation system.
    ZHANG Lisi, born in 1998, M. S. candidate. Her research interests include financial forecasting.
    MA He, born in 1997, M. S. candidate. His research interests include entity alignment.
  • Supported by:
    Ningxia Natural Science Foundation(2023AAC03316);Graduate Innovation Project of North Minzu University(YCX23146)

图谱嵌入传播的推荐方法

周北京1, 王海荣1,2(), 王怡梦1, 张丽丝1, 马赫1   

  1. 1.北方民族大学 计算科学与工程学院,银川 750021
    2.图像图形智能处理国家民委重点实验室(北方民族大学),银川 750021
  • 通讯作者: 王海荣
  • 作者简介:周北京(1997—),男,湖南衡阳人,硕士研究生,CCF会员,主要研究方向:基于知识图谱的推荐系统
    王海荣(1977—),女,宁夏石嘴山人,副教授,博士,CCF会员,主要研究方向:大数据知识工程、智能信息处理 bmdwhr@163.com
    王怡梦(2000—),女,河南三门峡人,硕士研究生,主要研究方向:多模态推荐系统
    张丽丝(1998—),女,云南曲靖人,硕士研究生,主要研究方向:金融预测
    马赫(1997—),男(回族),内蒙古赤峰人,硕士研究生,主要研究方向:实体对齐。
  • 基金资助:
    宁夏自然科学基金资助项目(2023AAC03316);北方民族大学研究生创新项目(YCX23146)

Abstract:

According to the richness of user and item information in Knowledge Graph (KG), the existing recommendation methods with graph embedding propagation can be summarized into three categories: user embedding propagation, item embedding propagation, and hybrid embedding propagation. The user embedding propagation method focuses on using items interacted with users and KG to learn user representations; the item embedding propagation method uses entities in KG to represent items; the hybrid embedding propagation method integrates user-item interaction information and KG, addressing the issue of insufficient information utilization in the first two methods. The technical characteristics of these three methods were deeply compared by specifically analyzing the key technologies of the three core tasks in the recommendation methods with graph embedding propagation: graph construction, embedding propagation, and prediction. At the same time, by replicating mainstream models in each category of methods on general datasets such as MovieLens, Booking-Crossing, and Last.FM, and comparing their effects using the CTR (Click-Through Rate) metric, it is found that the recommendation method with hybrid embedding propagation has the best recommendation performance. It combines the advantages of user and item embedding propagation methods, utilizing interaction information and KG to enhance the representations of both users and items. Additionally, a comparative analysis of various categories of methods was performed, their advantages and disadvantages were elaborated, and the future research work was also proposed.

Key words: Recommendation System (RS), Knowledge Graph (KG), Collaborative Filtering (CF), embedding propagation, Graph Neural Network (GNN)

摘要:

根据知识图谱(KG)丰富用户和项目信息的侧重不同,现有的图谱嵌入传播的推荐方法可归纳为用户嵌入传播、项目嵌入传播和混合嵌入传播这3类。用户嵌入传播方法侧重使用用户交互的项目和KG学习用户表示;项目嵌入传播方法使用KG中的实体表征项目;而混合嵌入传播方法融合了用户-项目交互信息和KG,以弥补前两类方法存在的信息利用不充分的不足。为深入对比3类方法的技术特点,重点剖析图谱嵌入传播的推荐方法中的图谱构建、嵌入传播和预测这3个核心任务的关键技术;同时,在MovieLens、Booking-Crossing和Last.FM通用数据集上复现每类方法中的主流模型,通过使用点击率(CTR)指标对比分析上述方法的效果。分析实验结果可知,混合嵌入传播方法的推荐性能最优,它综合了用户和项目嵌入传播方法的优势,利用交互信息和KG增强用户和项目表示;此外,对比分析每类方法,阐述各自的优缺点并展望未来的研究工作。

关键词: 推荐系统, 知识图谱, 协同过滤, 嵌入传播, 图神经网络

CLC Number: