基于贝叶斯网络的图书馆文献精准检索算法  

Accurate Retrieval Algorithm of Library Documents Based on Bayesian Network

作　　者：孙琳 金玉琴 王天一 SUN Lin;JIN Yu-qin;WANG Tian-yi(Information Division,Nanjing Stomatological Hospital,Medical School of Nanjing University,Nanjing 210008 China;Department of Orthodontics,Nanjing Stomatological Hospital,Medical School of Nanjing University,Nanjing 210008 China;Procurement Center,Nanjing Stomatological Hospital,Medical School of Nanjing University,Nanjing 210008 China)

机构地区：[1]南京大学医学院附属口腔医院信息科,江苏南京210008 [2]南京大学医学院附属口腔医院口腔正畸科,江苏南京210008 [3]南京大学医学院附属口腔医院采购中心,江苏南京210008

出　　处：《自动化技术与应用》2025年第4期57-60,98,共5页Techniques of Automation and Applications

基　　金：国家自然科学基金项目(81700939);南京市卫生科技发展项目(GAX22284)。

摘　　要：为提升图书馆文献检索效果,研究一种基于贝叶斯网络的图书馆文献精准检索算法。采用多分量模式设定检索术语,再计算权重,然后基于贝叶斯网络定性与定量分析图书馆文献,依据节点的相对独立性建立检索模型,获取检索节点与文献节点在概念空间中的覆盖程度。将网格调度器应用在模型中划分检索任务,通过贝叶斯网络的有向无循环图处理依赖关系,根据检索术语的执行时间确定文献调度顺序,实现图书馆文献的精准检索。实验测试结果表示,所提方法应用下能够获取较高的查全率和查准率,结果接近于99.8%,具有应用效果。In order to improve the effect of library document retrieval,this paper studies an accurate retrieval algorithm of library document based on Bayesian network.The multi-component model is used to set the retrieval terms,and then the weight is calculated.Then the library documents are analyzed qualitatively and quantitatively based on the Bayesian network,and the retrieval model is es-tablished based on the relative independence of the nodes to obtain the coverage of the retrieval nodes and document nodes in the concept space.The grid scheduler is applied to divide the retrieval tasks in the model,and the dependency is processed through the directed acyclic graph of Bayesian network,and the document scheduling order is determined according to the execution time of the retrieval terms,so as to realize the accurate retrieval of library documents.The experimental results show that the method in this paper can obtain high recall and precision,and the result is close to 99.8%,which has application effect.

关 键 词：贝叶斯网络 图书馆文献 精准检索 

分 类 号：TP391.3[自动化与计算机技术—计算机应用技术]