基于参考度的有效关联规则挖掘

摘要

People find that generated association rules may be false or redundant when using the traditional evaluating indicator. The article introduces the new measure-consult to improve measure system of mining algorithm. The experiments show that the efficiency of valid rules can be enhanced by adding the new measure.%针对当前关联规则挖掘采用的支持度-置信度框架在具体应用中存在的问题,引入新的指标--参考度对模型中关联规则挖掘算法评价体系进行改进.实验表明,通过引入参考度,能够提高挖掘有效规则的效率.