高校读者借阅行为的关联分析及应用实践

Correlation Analysis and Application of University Reader Borrowing Behavior

[目的/意义]旨在为提升高校图书馆的个性化推荐能力提供参考。[方法/过程]梳理了APRIORI关联分析在图书馆领域的理论研究和应用现状;结合图书借阅特征,对基于学科细分读者群体和分类号优化2个方面展开关联分析,对图书流通数据进行采集和清洗;借助Clementine可视化平台,利用"网络图"节点和APRIORI模型进行图书分类号的链接分析和关联规则建模,并进行评估;阐述如何将关联规则应用到实际的图书推荐服务。[结果/结论]基于Clementine可视化平台展开的读者借阅行为关联分析为高校图书馆提升馆藏建设质量和精准化服务提供了参考。

[Purpose/significance]The paper aims to provide references for improving the personalized recommendation ability of university libraries.[Method/process]The paper summarizes the theory research and application status of the APRIORI correlation analysis,combines with the feature of library,puts forward the correlation analysis based on subject subdivision readership and classification optimization.Then,the paper collects and cleans the book circulation data,and utilizes the network diagram node and APRIORI model with Clementine visualization platform,in order to conduct book classification number link analysis,model association rules and evaluate the results.Finally,the paper expounds how to apply association rule to the actual book recommendation service in detail.[Result/conclusion]The correlation analysis of reader borrowing behavior based on Clementine visualization platform can provide useful references for university libraries to improve the quality of collection construction and accurate service.