摘要
文章阐述知识图谱加持的智慧图书馆推荐技术体系构建基础,按推荐处理流程描述知识图谱加持的智慧图书馆推荐技术体系,据此初步构建基于图书资源知识图谱的智慧图书馆推荐系统,以验证所构建技术体系可行性。知识图谱加持的智慧图书馆推荐技术体系按推荐处理流程及其所用技术分为数据采集及预处理、基于Lambda架构的数据分析、馆藏知识图谱构建、特征建模、推荐生成、推荐结果展示6类。依托高性能大数据技术及并行化框架高效处理全域、实时交互数据,基于知识图谱聚合关联、挖掘发现语义知识以扩展组合推荐特征,可视化节点及关系以表征、解释推荐结果,最终提升推荐系统智能性、优化读者服务体验及赋能知识组织策略。
In this article,the authors expound the foundation for building smart library recommendation technology system supported by knowledge graph,give a description of the system according to recommendation processing flow,and develop a preliminary system for smart library recommendation business based on knowledge graph of book collections to verify the feasibility of the constructed technology system.The recommendation technology system of smart library supported by knowledge graph can be divided into six categories in accordance with recommendation processing flow and the technology used,which include data collection and preprocessing,data analysis based on Lambda architecture,collection-based knowledge graph development,feature modeling,recommendation generation,and recommendation result display.Relying on high-performance big data technology and parallelization framework,it can efficiently process all domain acquisition and real-time data interaction.Based on knowledge graph,it presents the aggregation and association of data,and the mining and discovery of semantic knowledge,so as to expand and combine the recommendation features.Relying on the visualization of nodes and relationships,it represents and explains the recommendation results,and thus ultimately improving the intelligence of recommendation system,optimizing readers’service experience and enabling knowledge organization strategies.
作者
孙雨生
祝博
范颖
SUN Yusheng;ZHU Bo;FAN Ying
出处
《图书馆论坛》
北大核心
2023年第10期77-87,共11页
Library Tribune
基金
教育部人文社会科学研究规划基金项目“基于本体的数字图书馆语义用户兴趣模型构建机理及应用模式研究”(项目编号:17YJA870016)
全国教育科学规划基金项目“基于多数据源、多方法融合的学科知识图谱构建方法研究”(项目编号:DIA160326)
湖北省社会科学基金项目“语义网格环境下数字图书馆个性化推荐机制及其应用研究”(项目编号:2015108)研究成果。
