期刊名称：International Journal of Advanced Research in Computer Engineering & Technology (IJARCET)
印刷版ISSN：2278-1323
出版年度：2018
卷号：7
期号：3
页码：298-302
出版社：Shri Pannalal Research Institute of Technolgy
摘要：Association rule mining deals with extracting important information hidden in large volumes of data and the standard metrics in the rule shows the relationship among the itemsets. There are different approaches which can be used to find interesting patterns or frequent patterns from a large pool of data. Support and Confidence are the standard metrics to measure the interestingness of the generated patterns. The tradition methods of evaluating these two standard metrics sometimes do not work well when the collected pool of data contains errors or noise in it. This paper proposed an approach for evaluating values of standard metrics of the frequent patterns generated from the pool of data which is applicable to the noise tolerant itemsets.
