摘要
提出了一种新颖的普遍化关联规则挖掘算法GARL。该算法连续扫描数据库事务序列，在最多不超过两遍扫描后生成所有频繁项目集，在首次扫描数据库时，能为用户给出反馈信息，允许用户对最小支持率进行调整，该算法能连续处理事务序列，可用于网上在线数据挖掘。
A novel algorithm GARL for mining generalized association rules is proposed. It continuously scans transaction sequences in database and produces a set of all frequent itemsets for a user-specified m inimum support after at most two scans. During the first scan of the database, it can give continuous feedback and allows user to change the minimum support. GARL processes a transaction sequences continuously and can be used for on-line data mining on network.
出处
《计算机工程》
CAS
CSCD
北大核心
2003年第7期4-6,共3页
Computer Engineering
基金
国家自然科学基金(60173058)
陕西省教育厅科学研究基金(00JK021)
关键词
知识发现
数据挖掘
关联规则
普遍化关联规则
Knowledge discovery
Data mining
Association rules
Generalized association rules
