本文在数据挖掘的过程中,以我馆金盘图书管理系统中流通库的借阅记录为样本,利用Clementine的两步聚类模型进行数据挖掘,将读者聚类为惰性型、一般型和活跃型三种类型,实现对读者的聚类分析,运用数据挖掘结果来主动发现读者的需求,以提供图书馆在图书推荐和图书馆决策等工作时作为客观的参考依据。 In the process of data mining, taking the loan record of the circulation library in Gold Library Management System of our library as a sample, we use Clementine’s two-step clustering model for data mining, and classify the readers as inert type, general type and active type Three types of clustering analysis to achieve the readers, the use of data mining results to proactively detect the needs of readers to provide libraries in the book recommendation and library decision work as an objective reference.