知识图谱与推荐算法任务优化及领域应用发展综述OA

As the core technology of current information filtering,recommendation algorithms need to solve the problems of information fragmentation and insufficient interpretability through semantic association of knowledge graphs and struc-tured knowledge.In order to further promote the technological integration of knowledge graph and recommendation algo-rithm,research the key to improve the performance of recommendation models and the development of multi-domain and multi-technology,this paper decomposes the process of combining knowledge graph with recommendation algorithm into three parts:knowledge graph module optimization,recommendation task module optimization,and multi-task module optimization.The paper comprehensively reviews the relevant theories and research results of each module from the per-spectives of development history,optimization and improvement,core technology,etc.Simultaneously,catering to the current development trends in the field,the latest developments and challenges in the applications of large language models,medicine and e-commerce fields are summarized.Finally,a summary and outlook is provided on the existing problems of knowledge graphs and recommendation algorithms,providing new ideas for future research.