考虑用户行为的图书馆文献个性化检索算法设计  

Design of Personalized Retrieval Algorithm for Library Documents Considering User Behavior

作　　者：王瑞[1] 李爽[2] WANG Rui;LI Shuang(Library,Dalian Naval Academy,Dalian 116000 China;Library,University of Science and Technology,Anshan 114000 China)

机构地区：[1]海军大连舰艇学院图书馆,辽宁大连116000 [2]辽宁科技大学图书馆,辽宁鞍山114000

出　　处：《自动化技术与应用》2024年第11期128-131,共4页Techniques of Automation and Applications

基　　金：辽宁省图书馆学会研究项目(2021tsgxhqnkt-011)。

摘　　要：在海量的文献检索过程中,用户行为这一重要检索约束条件被抛弃,仅仅以检索数据特征作为推导,导致检索准确性下降。设计一种考虑用户行为的图书馆文献个性化检索算法。首先设计算法的总体逻辑框架,建立用户行为模型获取邻居节点信息,计算权重值,获取不同用户行为的相似度系数。以行为相似度系数作为约束条件,设计多层次目标个性化文献索引算法,在随机游走模型下获取个性化文献检索模型的优化目标,并得到文献检索结果。实验结果显示:该检索算法最低可在18 ms的时间内完成对应文献检索,文献检索精度均在90%以上,且文献检索存储开销等测试,均能够得到较好的结果,可见该算法性能良好。In the process of massive document retrieval,user behavior,an important constraint of retrieval,is abandoned,and only the retrieval data characteristics are used as the derivation,resulting in the decline of retrieval accuracy.It designs a personalized retrieval algorithm for library documents considering user behavior.Firstly,the overall logical framework of the algorithm is designed,and the user behavior model is established to obtain the neighbor node information,calculate the weight value,and obtain the similarity coefficient of different user behaviors.Taking the behavioral similarity coefficient as the constraint,a multi-level objective personalized document index algorithm is designed to obtain the optimization objectives of the personalized document retrieval model under the random walk model,and obtain the document retrieval results.The experimental results show that the retrieval algorithm can complete the corresponding document retrieval in at least 18 ms,the document retrieval accuracy is more than 90%,and the document retrieval storage cost and other tests can get good results,which shows that the algorithm has good performance.

关 键 词：个性化检索 检索算法 用户行为模型 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程] TP391.3[自动化与计算机技术—控制科学与工程]