摘要
在挖掘关联规则过程中,用户往往需要多次调整(增加或减少)最小支持度,才能获得有用的关联规则。给出一个利用已存信息有效产生新候选项目集的PSI算法,结果表明每次扫描数据库时能有效地减少候选项目集的数。
In mining association rules, an user may required to tune (increase or reduce) the value of the minimum support many times before a set of useful association rules could be obtained from the transaction database. In this paper, an algorithm PSI was given for efficient generation of new candidate itemsets using prestored information. It can significantly reduce the number of candidate itemsets in each database scan.
出处
《计算机应用》
CSCD
北大核心
2005年第6期1360-1361,1372,共3页
journal of Computer Applications
基金
云南省教育厅科研基金资助项目(022Y165)