摘要
提出了一种有效的多维多层关联规则增量更新算法(EARIU).对商品信息进行概念分层,把用户属性作为维信息,放入FP树中进行挖掘,使挖掘得到的关联规则有更多有趣信息.根据购买数量对支持度阈值进行调整,利用新增记录对原有项/维集的影响的本质特性,更新原有和挖掘新增数据库的频繁项/维集.
An effective multi-dimension and multi-level association rules incremental update algorithm(EARIU)is proposed in this article.A conceptual hierarchy tree is built from the database and commodity information,with feature selection for dimension information,to mine multi-dimension and multi-level association rules from the tree.More interesting information can be acquired.Adjusting the supporting threshold by considering the purchase number and utilizing the co-relation between the incremental record and the ...
出处
《微电子学与计算机》
CSCD
北大核心
2009年第3期113-116,共4页
Microelectronics & Computer
基金
天津工程师范学院科研发展基金(SK-JH2007001)
关键词
关联规则
增量更新
商品概念层次
频繁项/维集
association rule
incremental updating
product catalog
frequent item sets and dimension sets
