摘要
[目的/意义]研究如何将数据挖掘技术应用于图书采购决策。[方法/过程]以图书的有效借阅数为基础,以具有预测性质的决策指标——借阅热度为研究对象,运用决策树分类技术对历史借阅记录进行挖掘。[结果/结论]构建具有预测能力的模型,采用C4.5决策树算法预测候选采购图书借阅热度值,并进行实验验证。实验结果表明,决策树分类技术能够较为准确地预测候选图书借阅热度。
[Purpose/significance]The paper is to discuss how to apply data mining into book acquisition decision. [Method/process]The paper bases on effective number of book borrowing, takes borrowing heat, a predictable decision indicator, as research object,and uses decision tree to mine historical borrowing records. [Result/conclusion]The paper constructs a predictable model, adopts C4.5 decision tree algorithm to predict candidate book＇s borrowing heat, and conducts an experimental verification. The results show that decision tree classification can predict candidate book＇s borrowing heat accurately.
出处
《情报探索》
2016年第10期76-79,共4页
Information Research
关键词
图书采购
决策指标
数据挖掘
图书分类
决策树
book acquisition
decision indicator
data mining
book classification
decision tree
