Improving Efficiency of META Algorithm Using Record Reduction

Erasable Itemset Mining is the key approach of data mining in production planning. The erasable itemset mining is the process of finding erasable itemsets that satisfy the constraint i.e. user defined threshold. Efficient algorithm to mine erasable itemsets is extremely important in data mining. Since the META Algorithm was proposed to generate the erasable itemsets. In last few years there have been several methods to improve its performance. But they do not consider the time constraint. If database is large META takes too much time to scan the database. In this paper, Author purposed an Improved META (I-META) algorithm which reduces the scanning time by reduction of production records. It also reduces the redundant generation of sub-items during trimming the candidate itemsets, which can find directly the set of erasable itemsets and removing candidate having a subset that is not erasable.