摘要
关联规则是数据挖掘中的重要研究内容之一,国内现有的关联规则算法大多是研究挖掘数据库不变的限定条件下,发现挖掘数据的各属性间的所有关联型知识.而事实上大多数挖掘数据会随时间的变化不断变化.针对数据库中追加数据时,如何有效地更新关联规则的问题,提出了一种新算法———IUAMAR算法.该算法可以有效地利用知识数据库中保留的最小非高频繁项目集产生新的候选项目集,避免了候选项目集的数量太庞大的问题.
Mining of the association rules is an important issue in the data mining field. Al present most of algorithms for mining association rules researched are restricted in unchanged mining database inland,which can find all association knowledge among data attribute by mining . But in lact most of mining data can change continuously with time. This paper proposed a new 'algorithm - - IUAMAR algorithm in order to npdate association rules when mining database are increasing. This algorithm generate new eandidate itemsets by using effectively minimum infrequent itemsets in knowledge dalabase and avoid generating a large number of candidate itemsets.
出处
《湘潭大学自然科学学报》
CAS
CSCD
北大核心
2006年第3期36-39,共4页
Natural Science Journal of Xiangtan University
基金
湖南省教育厅项目资助项目(05C574)
关键词
关联规则
最小非高频敏项目集
增量更新
IUAMAR
Association rules
Minimum infrequent itemsets
Incremental updating
Incremental Updataing Algorithm for Mining Association Rules