基于商空间理论多维多层次关联规则挖掘算法研究

Research of Multi-Dimension Multi-Level Association Rules Mining Algorithm Based on Theory of Quotient Space

传统的FP-growth算法在OLAP多维多层次关联规则挖掘过程中,存在明显不足,提出了基于商空间理论的多维多层次关联规则挖掘QCHFP-growth算法。在对其进行多维多层次频繁项集挖掘性能测试时,实验结果表明,较FP-growth算法在挖掘效率方面有显著提高。

As the shortage of traditional FP-growth algorithm mining association rules of multi-dimension multi-level, this paper introduces an algorithm of Quotient Concept Hierarchies Based on FP-growth （QCHFP-growth）. The experimental results show that the QCHFP-growth algorithm has the better performance compared with the FP-growth algorithm in mining frequent itemset of multi-dimen- sion multi-level.