Mining recent frequent itemsets in data streams by radioactively attenuating strategy

Abstract

We propose a novel approach for mining recent frequent itemsets. The approach has three key contributions. First, it is a single-scan algorithm which utilizes the special property of suffix-trees to guarantee that all frequent itemsets are mined. During the phase of itemset growth it is unnecessary to traverse the suffix-trees which are the data structure for storing the summary information of data. Second, our algorithm adopts a novel method for itemset growth which includes two special kinds of itemset growth operations to avoid generating any candidate itemset. Third, we devise a new regressive strategy from the attenuating phenomenon of radioelement in nature, and apply it into the algorithm to distinguish the influence of latest transactions from that of obsolete transactions. We conduct detailed experiments to evaluate the algorithm. It confirms that the new method has an excellent scalability and the performance illustrates better quality and efficiency. © Springer-Verlag Berlin Heidelberg 2005.