数据驱动的高校图书馆用户留存研究  

Data-driven Study on User Retention in University Libraries

作　　者：胡婧 陈添源 詹庆东[2,3] 林艺山 HU Jing;CHEN Tianyuan;ZHAN Qingdong;LIN Yishan

机构地区：[1]闽南师范大学图书馆 [2]福州大学图书馆 [3]福建省高校图工委

出　　处：《图书馆论坛》2022年第11期69-77,共9页Library Tribune

基　　金：教育部人文社科基金项目“基于用户画像的高校图书馆精准服务模式构建及实证研究”(项目编号:20YJC870002)研究成果。

摘　　要：数据驱动的用户留存分析能为高校图书馆深化个性化服务提供精准化的决策支持。文章以高校图书馆业务战略目标为出发点,探索构建用户留存分析框架和方法体系。选取具备全量数据特征的文献提供服务平台为实证对象,对接业务系统后台采集、清洗和存储用户留存行为全量数据,从用户生命周期分析、服务复用周期特征挖掘、结合NES理论的同期群分析获取用户留存区间特征分布及其留存率,重构RFM模型形成用户留存行为标签体系RFCLS,继而通过KMeans聚类获取和描述用户拉新群、选择留存群、忠诚用户群和易流失用户群等分群的用户画像,并据此提出相适宜的精准化营销服务策略。Data-driven study on user retention can provide accurate decision support for university libraries’deepening personalized services.Taking the strategic goals of university libraries as the starting point,this paper attempts to establish a framework as well as a methodological system of user retention analysis.Taking those literature service platforms with full data characteristics as the empirical object,it first collects all data related with the collecting,cleaning and storing of user retention behaviors.Then,by means of user life cycle analysis,service reuse cycle feature mining and simultaneous group analysis combined with NES theory,it attempts to obtain the feature distribution and retention rate of user retention interval,reconstruct the RFM model,and form the user retention behavior label system RFCLS.Besides,by means of KMeans cluster,it tries to obtain and describe the persona of user groups,such as new user groups,selected retention groups,loyal user groups and vulnerable user groups.In the end,it proposed some strategies for providing more accurate marketing services.

关 键 词：用户留存 数据驱动 同期群分析 资源服务 

分 类 号：G258.6[文化科学—图书馆学] G252