一种有效的基于约束的关联规则发现算法

基于约束的关联规则挖掘是针对特定约束的规则的挖掘，挖掘的结果有着更好的针对性和实用性，Separate算法是现有的效果较好的算法，但有2点不足：未修剪生成的候选集和候选项重复生成。对此该文提出了改进的SeparateP算法，算法中加入了对候选集的修剪，并且利用了项集重复生成的信息，使候选集的修剪更加有效快捷。实验表明，改进算法显著提高了原算法的效率。

Effective Algorithm for Mining Constrained Association Rules

Mining constrained association rules mines some special constrained rules and the results are more pertaining and practical. Separate Algorithm is a good algorithm to mine constrained association rules and has two main shortages. One is that it scans candidate itemsets without pruning; the other is that it creates candidate items redundantly. An improved algorithm, Separate P, is presented. The pruning on candidate sets is added to the algorithm, and Separate P makes full use of the information that the itemsets create repeatedly. The pruning is made more efficient and faster. The experimental results show that Separate P can increase the Separate efficiency.