近几年来,不确定数据广泛出现在传感器网络、Web应用等领域中。不确定数据挖掘已经成为了新的研究热点,主要包括聚类、分类、频繁项集挖掘、孤立点检测等方面,其中频繁项集挖掘是重点研究的问题之一。综述了传统的频繁项集挖掘的两类基本算法,分析了在此基础上提出的适用于不确定数据以及不确定数据流的频繁项集挖掘的方法,并探讨了今后可能的研究方向。
Uncertain data is widespread in some application fields such as sensor network,Web applications and so on.Uncertain data mining has become a new hotspot.Uncertain data mining includes clustering, classification, frequent itemsets mining, outlier detection, etc., in which frequent itemsets mining is one of the focus issues.This paper introduces two kinds of basic algorithms of mining frequent itemsets from traditional data: Apriori algorithm and FP-growth algorithm, and then analyses the methods proposed for mining frequent itemsets from uncertain data and uncertain data streams,A summary of research direction on uncertain data frequent itemsets mining is given.