Mining erasable itemsets

Abstract

In this paper, we introduce a new kind of mining problem — - mining erasable itemsets, which is de-rived from planning products of the manufacturing industry. For this problem, we first present the formal definition of mining erasable itemsets and discuss some basic properties of the problem. Based on our analyses, we propose an efficient algorithm, META, for mining erasable itemsets. For evaluating META algorithm, we built three synthetic product databases. The results of applying META algorithm to these synthetic databases show its effectiveness.