摘要
针对图书馆服务自动化感知效果不佳,导致无法实现学生个性化推荐的问题,提出将最近邻搜索K-means聚类算法与关联规则算法相结合,构建一个基于大数据挖掘技术的图书馆服务自动化感知模型。首先,基于读者借阅行为,采用最近邻搜索K-means聚类算法(NNSK-means)分别从个体层次和集体层次进行聚类分析,挖掘出读者的阅读趋势和阅读兴趣;然后通过基于Apriori关联规则算法进行图书间与院系间关联规则挖掘;最后基于挖掘结果进行读者借阅不同种类图书概率反映和分析,从而提升自动化感知和个性化推荐效果。结果表明,采用提出的基于大数据挖掘技术的图书馆服务自动化感知模型后,热门图书推荐服务与学生个性化推荐服务的准确率和学生满意度分别保持在98%和90%以上,表明构建的模型可实现图书馆服务自动化感知,模型性能良好,可为学生提供更加精准的个性化推荐服务。
In view of the problem that the automatic perception of library service is not effective and it is impossible to realize the personalized recommendation of students,the nearest neighbor search K-means clustering algorithm with the association rule algorithm is proposed to build an automatic perception model of library service based on big data mining technology.Firstly,based on the reader borrowing behavior,the nearest neighbor search K-means clustering algorithm(NNSK-means)is used to analyze at the individual and collective level to dig out the reading trend and interest;then dig between the books and departments based on Apriori association rules and analyze the probability of different kinds of books based on the mining results,so as to improve the automatic perception and personalized recommendation effect.The results show that the library service based on big data mining technology automation perception model,popular book recommendation service and students personalized recommendation service accuracy and student satisfaction remain at 98%and 90%respectively,shows that the building model can realize the library service automation perception,model has a good performance,can provide students with more accurate personalized recommendation service.
作者
张婷
ZHANG Ting(Xinjiang University,Urumqi 830046,China)
出处
《自动化与仪器仪表》
2023年第7期5-9,共5页
Automation & Instrumentation
基金
2021年新疆高等学校图书情报工作委员会科研项目《新疆高校图书馆服务创新研究》(TGW-2021022)。
