摘要
[目的/意义]跨域推荐通过挖掘、迁移并融合利用不同来源的知识为用户提供个性化的推荐服务,近年来得到学术界和工业界的热点关注,文章从多个角度对跨域推荐中的知识融合研究进展进行梳理和归总。[方法/过程]首先对跨域推荐问题进行了系统地分析,探讨了跨域推荐的“域”、跨域推荐场景和跨域推荐任务中的知识需求;其次对跨域推荐中的知识融合方法进行了分类,总结了基于聚类、基于语义、基于图模型和基于标签关联的跨域知识融合方法的优点和不足;然后分析了知识图谱技术对于跨域知识融合的启发;最后对跨域推荐研究中的知识融合进行了总结和展望。[结果/结论]知识融合作为跨域推荐研究中的重要环节,可以在知识层面实现对数据的深度加工和有效利用,为跨域推荐问题研究与实践提供新的范式。
[Purpose/Significance]Cross Domain Recommendation provides personalized services for users by mining, migrating, and integrating preference knowledge from different sources, which has attracted the attention of academia and industry in recent years.This paper summarizes the research progress of knowledge fusion in cross domain recommendation from multiple perspectives.[Method/Process]Firstly, this paper systematically analyzed the problem of cross domain recommendation, and discussed the domain of Cross Domain Recommendation, the CDR task, and the knowledge requirements in the CDR task;Secondly, the knowledge fusion methods in CDR were classified, and the advantages and disadvantages of cross domain knowledge fusion methods based on clustering, semantics, graph model and label association were summarized.The inspiration of knowledge graph technology for cross domain knowledge fusion was analyzed, the challenges and development of Cross Domain Recommendation knowledge fusion were discussed.Finally, the knowledge fusion in cross domain recommendation research was summarized.[Result/Conclusion]Knowledge Fusion is an important part of cross domain recommendation research.It can realize the deep processing of data at the knowledge level, and provide a new paradigm for the research and practice of cross domain recommendation.
作者
张彬
徐建民
吴姣
Zhang Bin;Xu Jianmin;Wu Jiao(School of Management,Hebei University,Baoding 071002,China;Magazine House,Hebei University,Baoding 071002,China)
出处
《现代情报》
CSSCI
2023年第3期157-166,共10页
Journal of Modern Information
基金
河北省社会科学基金项目“基于知识图谱的查询扩展研究”(项目编号:HB21TQ005)
中国高校产学研创新基金—新一代信息技术创新项目“基于知识图谱的信息检索教学平台设计研究”(项目编号:2020ITA02045)
河北省社会科学发展研究课题“媒体融合视域下河北省社科学术期刊影响力提升研究”(项目编号:20220202057)。
关键词
跨域推荐
知识融合
个性化推荐
知识图谱
cross domain recommendation
knowledge fusion
personalized recommendation
knowledge graph
