基于候选项集个数上阶的增量式关联规则更新算法

A General Incremental Algorithm for Mining Association Rules

提出了一种有效的增量式关联规则挖掘算法IAR ,算法的特点在于 :提出并采用了基于候选项集个数上阶的选择扫描数据库的机制 ,可有效减少数据库的扫描次数 ;算法是一种通用的增量式算法 ,提出了最小支持度和数据库均改变时 ,增量式挖掘中的重要性质 ,从而可充分利用上一次挖掘的结果 ,有效减少候选项集的数目 .并且提出了基于组合数学和项集等价类理论的计算候选项集个数的上阶的方法 .通过大量的数据实验 ,表明算法的效率比已有的算法有了很大提高 .

Mining of association rules is one of the most important fields in data mining.In this paper,a new general incremental algorithm IAR for mining association rules is presented.The disting uishing feature of IAR is as follows:First,the selective scan strategy is adopted,which is based on the upper bound on the number of candidate itemsets;Second,IAR can efficiently update the discovered rules when the value of support threshold is changed,new transactions are added to the database,and obsolete ones are removed from it.Furthermore,based on the Kruskal-Katona theorem and the itemset equivalence class theory,we devise the technique for calculating the maximal number of candidate itemsets.Experiments show the prominent performance of the algorithm IAR.