摘要
间接关联是数据挖掘领域中一种数据项之间的关联关系,可有效地应用于市场营销及Web日志分析等领域.现有的间接关联挖掘算法采用Apriori算法框架,需挖掘出所有的频繁项目集,因而存在挖掘效率低的缺陷.为此,提出了一种基于前缀广义表的快速间接关联挖掘算法,该算法无须生成所有的频繁项目集且仅须扫描数据库2遍,可有效提高间接关联的挖掘效率.
Indirect association is an associated relationship between items and frequent itemsets in data sets. There are many potential applications for indirect associations, such as database marketing and web - log analysis, etc. Existing algorithms need to generate all frequent itemsets using Apriori - like framework. Hence, they are in low efficiency. This paper proposes an algorithm based on prefix general list for mining indirect associations-ABPGLMIA, which improves the mining efficiency of indirect associations by scanning database twice. Experimental results show that the algorithm ABPGLMIA is efficient.
基金
安徽省自然科学基金资助项目(03042205)安徽省教育厅教学研究基金资助项目(2003kj029)
