基于维分类的关联规则的元规则制导挖掘

摘要

元规则制导的关联规则挖掘可以提高挖掘过程的效率和精确度,目前已经提出了许多关联规则的元规则制导挖掘算法,尤其是在关系数据库中；而在数据立方体上的元规则制导挖掘算法相对较少,且大多数是基于Apriori思想的算法,它们都存在冗余谓词搜索的问题.针对这种情况,提出了一种以元规则中维度的不同类型为依据的改进算法LRS,并在实验中证明了算法的有效性.%Metarule-guided mining of association rules can improve the efficiency and accuracy of mining process.Currently many algorithms have been proposed, especially in relational databases.In the case of data cubes, many algorithms are proposed, and most of them are based on Apriori theory, which lead to the problem of redundant predicates searching.In view of this situation,this paper proposes an algorithm LRS according to different types of dimensions in the metarule,and the experiment results demonstrate the effectiveness of the proposed algorithm.