摘要
针对挖掘稠密的长模式关联规则时,需要大量的存储空间、求长集合的交集时需要大量的计算时间以及计算候选频繁项集的支持度时需要访问反复扫描数据库,提出了基于信息表和差集的挖掘算法。实验证明,在相同的数据库和支持度情况下,该算法较apriorTID算法能减少挖掘时间和占用的空间。
When mining dense and long mode association rules,it needs massive storage space. When computing the intersection of long sets,it consumes much CPU time,and when computing candidate frequent itemset support,it needs to access and to repeatedly scan database. In light of these,a mining algorithm based on information table and difference set is put forward. Experiment proves that it can reduce the mining time and storage space compared to apriorTID algorithm under the circumstances of same database and support.
出处
《计算机应用与软件》
CSCD
2010年第12期202-204,共3页
Computer Applications and Software
关键词
关联规则
频繁项集
信息表
差集
Association rules Frequent itemset Information table Difference set