基于频繁集的多层次交互式关联规则挖掘

Multiple-Level Interactive Association Rules Mining Based on Frequent Sets

定义了一系列关联规则的相关概念 ,然后 ,针对现有的销售事务数据库 ,提出了一种改进的获取多层次信息方法 ,并据此对各数据进行压缩编码 ,减少处理时间 .同时 ,借助频繁集和交互式技术 ,实现多个概念层次交互式挖掘 ,以按用户实际需要提取出其感兴趣的关联规则 。

A series of notations relative to association rules was presented.According to the currenttrans- action database,a novel acquisition method of multiple- level taxonomy was proposed.Every transaction was encoded to reduce the processing time.With frequent itemsets and interactive techniques,the associa- tion rules that interest users were derived.The mining process and efficiency are greatly improved.