An Approach For Optimize Rule Mining Based on Domain Problem Information and Pareto Front

Having a look at growing amount of stored data on differentinformation systems such as shopping centers and medical data wecertainly need a method to make use of these stored data in the mostefficient way in order to help making decisions based on them. Butbecause of vast amount of data, we must choose some parts of them touse. In the most of time we decide on this data with two factor: supportand confidence but we don’t notice to another important factors likevolume of purchased items. When these factors grow up, a multi-valuedecision making method can be quite helpful. In this paper we putforward a new method for discovering association rules by notice tovolume and benefit of sold items, finally for have a good decisionbased on this rules a ranking approach which is based on pareto-front isgoing to be introduced, It is used to measure efficiency of associationrules for multi-goal purposes.

کلیدواژه ها:

Data mining ، Multi-goal decision making ، Pareto front ، Association rules