基于组合评价方法的关联规则兴趣度评价

Interestingness Evaluation of Association Rules Based on Combination Evaluation Method

关联规则挖掘算法通常生成大量的规则,但由于资源的限制,只有少量规则可能被筛选出来使用。因此关联规则的兴趣度评价成为数据挖掘领域中的一个重要问题。考虑到关联规则兴趣度评价本质上是一个多属性决策问题,本文首先基于关联规则的客观兴趣度度量和用户的主观偏好,建立了关联规则评价指标体系;然后提出一种基于组合评价方法的关联规则评价的框架及其具体实现步骤,以解决多种评价方法评价结果不一致的问题;最后以某超市购物篮数据分析为例,基于整体差异的组合评价方法实现了关联规则的组合评价以验证所提评价方法的可行性和有效性。

Numerous rules can be generated by association rule mining algorithms.But only a small number of these rules may be selected for implementation due to the limitations of resources.Accordingly,evaluating the interestingness of association rules becomes an important issue in data mining.Since the interestingness evaluation of association rules could be a multiple attributes decision problem essentially,in this paper an evaluation index system for association rules is built based on objective interestingness measures and the users＇ subjective preferences,and then a framework for association rules evaluation based on combination evaluation method,together with its implementation procedures is proposed,which could deal with the inconsistency problem existing among different evaluation methods.In the end,taking market basket analysis as an example,an objective combination evaluation method based on whole diversity maximization is applied to evaluate the association rules generated from a grocery database to illustrate the feasibility and effectiveness of the proposed method.