基于艾宾浩斯遗忘曲线的零售商品模糊关联分析

针对消费者对商品的偏好存在时序变化特征,而传统关联规则方法未考虑时间因素的影响,且对海量数据集进行关联挖掘时存在效率低下的问题,提出了基于艾宾浩斯遗忘曲线的模糊关联规则算法。该方法通过FCM聚类算法对商品进行聚类,并用艾宾浩斯遗忘曲线来修正聚类的距离度量方法,从而得到商品类及各类的代表点商品;然后将各代表点商品作为属性,消费记录小票作为样本,利用模糊关联规则算法得到代表点商品间的规则;最后将某大型超市一个月的销售记录作为关联规则的事务数据来挖掘潜在规律,结果显示所提算法先对商品模糊关联分析,与传统直接对商品进行关联分析相比,该算法可以显著提高关联挖掘的效率和规则的正确率。...

Full description

Saved in:
Bibliographic Details
Published in计算机应用研究 Vol. 35; no. 2; pp. 462 - 465
Main Author 李桃迎;张鑫;陈燕
Format Journal Article
LanguageChinese
Published 大连海事大学交通运输管理学院,辽宁大连,116026 2018
Subjects
关联规则
模糊关联规则
聚类分析
艾宾浩斯遗忘曲线
模糊关联规则
association rules
fuzzy association rules
艾宾浩斯遗忘曲线
cluster analysis
聚类分析
Ebbinghaus forgetting curve
关联规则
Online AccessGet full text
ISSN1001-3695
DOI10.3969/j.issn.1001-3695.2018.02.030

Cover

More Information
Summary:针对消费者对商品的偏好存在时序变化特征,而传统关联规则方法未考虑时间因素的影响,且对海量数据集进行关联挖掘时存在效率低下的问题,提出了基于艾宾浩斯遗忘曲线的模糊关联规则算法。该方法通过FCM聚类算法对商品进行聚类,并用艾宾浩斯遗忘曲线来修正聚类的距离度量方法,从而得到商品类及各类的代表点商品;然后将各代表点商品作为属性,消费记录小票作为样本,利用模糊关联规则算法得到代表点商品间的规则;最后将某大型超市一个月的销售记录作为关联规则的事务数据来挖掘潜在规律,结果显示所提算法先对商品模糊关联分析,与传统直接对商品进行关联分析相比,该算法可以显著提高关联挖掘的效率和规则的正确率。
Bibliography:51-1196/TP
Ebbinghaus forgetting curve; cluster analysis; association rules; fuzzy association rules
Consumers＇ preference for goods has the characteristics of sequential variation. However, the traditional associa- tion rule method has not considered the influence of time factor, and there is the problem of low efficiency in the association mining of massive data sets. Therefore, this paper proposed a fuzzy association rule algorithm based on Ebbinghaus forgetting curve. First of all, this algorithm employed the FCM clustering algorithm to cluster the goods, and introduced the Ebbinghaus forgetting curve for revising the clustering distance metrics, so that to get the goods clusters and all kinds of representative point goods. And then, it used each representative point as the attribute, and used the shopping receipt as the sample, and utilized fuzzy association rules algorithm to get the rules of the representative points. Finally, this paper used the sales records of a large supermarket as the transaction da
ISSN:1001-3695
DOI:10.3969/j.issn.1001-3695.2018.02.030