摘要
提出了一种基于品类聚类的关联规则优化算法。该算法首先根据文中定义的品类特征向量,用结构化的数据来表示事务;然后根据一种基于密度的聚类算法,对结构化的数据进行聚类,同时将对应的原始事务进行聚类;最后根据聚类后得到的类的长度以及用户指定的最小支持度,确定类内的最小支持度,在类内挖掘关联规则。实验结果表明,与传统算法相比,该算法效率较高,具有一定的实用价值。
A category-based algorithm is proposed in this paper to enhance ruining association rules. This algorithm involves three steps of processing：transactions are first represented as transaction-category points based on the predefined category feature vectors;these transactioncategory points and transactions are then clustered with the applied clustering algorithm ;finally, the minimum support of the cluster is calculated based on the result number of points in a specified cluster and the user-defined minimum support, and a traditional algorithm is applied to get association rules in the cluster. The experimental result shows that this algorithm is effective and practical.
出处
《计算机应用与软件》
CSCD
北大核心
2007年第1期140-142,184,共4页
Computer Applications and Software
关键词
品类信息
事务聚类
关联规则
Category information Transaction clustering Association rules
