摘要: 分析了Apriori核心算法,举例说明了其设计思想上的不足,并重新定义了关联规则形式和引进了兴趣度的概念。主要定义了合理的兴趣度,即基于可信度和支持度方差的兴趣度 = *( + ),并因此而设计了基于此兴趣度定义的关联规则挖掘算法,并对算法做了适当的分析。
关键词:
兴趣度,
关联规则挖掘算法,
关联规则,
可信度,
支持度
Abstract: Based on the analysis of Apriori algorithm, this paper points to the defects of its design thought. After adopting the concept of interest measure and new definition of association rules, it defines the more reasonable interest measure that bases on the difference between the square of confidence and the square of support, also = *( + ). Therefore, with the new definition of interest and association rules, it designs a new association rules mining algorithm and also gives some analysis on it.
Key words:
Interest measure,
Association rules mining algorithm,
Association rules,
Confidence degree,
Support degree
中图分类号:
马建庆;钟亦平;张世永. 基于兴趣度的关联规则挖掘算法[J]. 计算机工程, 2006, 32(17): 121-122,.
MA Jianqing;ZHONG Yiping;ZHANG Shiyong. Association Rules Mining Algorithm Based on Interest Measure[J]. Computer Engineering, 2006, 32(17): 121-122,.