Abstract: In the mining methods of association rules, itemset with high-length usually have lower support, but still have potential value. To efficiently mine association rules with long-pattern, a length-decreasing support constraint is proposed. This paper uses vector to mine association rules under this constraint. Compared to other mining methods of association rules, the new method can mine more efficacious long-pattern, reduce many unnecessary short-pattern and improve efficiency of mining association rules.

Key words: association rules, length-decreasing support, vector

摘要： 关联规则挖掘中的长模式，通常支持度较低，但仍然具有潜在的价值。为了挖掘长模式下的有效关联规则，该文提出了一种在新的长度递减支持度约束条件下采用向量法进行的关联规则挖掘。该方法能够挖掘更多有效的长模式，减少无用的短模式，提高了关联规则挖掘的效率。

关键词: 关联规则, 长度递减支持度, 向量法

CLC Number: 

• TP311