摘要
传统图书馆检索系统往往依赖关键词匹配,难以捕捉用户深层需求,导致信息检索效率低下,用户体验不佳。本研究分析了用户在图书馆检索服务的场景需求,设计总体思路及核心模块,融合读者多维度特征和知识图谱技术,将用户检索意图与知识图谱相结合,理解与识别用户意图并进行个性化推荐。该方法能精准解析检索意图,实现资源智能筛选与推荐,在检索资源,结果呈现与用户推荐上更有优势,为图书馆知识服务转型提供了有效路径。
Traditional library retrieval systems often rely on keyword matching,struggling to capture users’deep-seated needs,resulting in low efficiency and poor user experience.This study analyzes user scenario requirements in library retrieval services,designs the overall framework and core modules,and integrates multi-dimensional reader characteristics with knowledge graph technology.By combining user search intent with the knowledge graph,the system understands and identifies user intentions to deliver personalized recommendations.The proposed method accurately parses retrieval intentions,achieves intelligent resource filtering and dynamic recommendations,demonstrating superior performance in resource discovery,result presentation,and user-centric recommendations compared to traditional services.Experimental results show significant improvements in recall rate,response speed,and user satisfaction,providing an effective approach for transforming library knowledge services.
作者
刘飘飘
陈臣
Liu Piaopiao;Chen Chen
出处
《新世纪图书馆》
2025年第6期62-68,92,共8页
New Century Library
基金
2021年度甘肃省科技厅重点研发计划项目“平安甘肃大数据智慧调度与实时可视化展示研究”(项目编号:21YF5FA087)系列研究成果。
关键词
知识图谱
读者特征
智能检索
推荐系统
自然语言处理
Knowledge graph
Reader characteristics
Intelligent retrieval
Recommender systems
Natural language processing
