一种带兴趣度的正负关联规则挖掘算法

摘要

In terms of the negative association rules mining problems, maximum support is proposed to be used to control the growth of the frequent itemsets. The formula of correlation is improved and used as an interest measure for the positive and negative association rules to get rid of the uninterested rules. The former and back association rules are limited to ensure the practicability and comprehensibility of the association rules. At last, this paper gives the algorithm which can simultaneously mine positive and negative association rules. And the experimental results show that this algorithm is efficient and feasible.%针对负关联规则挖掘所带来的问题，提出加入最大支持度来控制频繁项集生成规模，改进了相关性的计算公式，并将其用作正负关联规则的兴趣度来剔除无兴趣的关联规则，限制关联规则中的前后件项目个数来保证挖掘出的关联规则的实用性和可理解性。最后，给出一种能够同时挖掘正负关联规则的算法，实验结果表明算法是有效的、可行的。