An efficient mining scheme for high utility itemsets

TL;DR

This paper introduces R-Miner, a new high utility itemset mining algorithm that uses residue maps to improve efficiency, outperforming existing algorithms in speed and memory usage.

Contribution

The paper presents R-Miner, a novel algorithm utilizing residue maps for faster and more memory-efficient high utility itemset mining.

Findings

R-Miner outperforms benchmark algorithms in execution time.

R-Miner uses residue maps to store utility information efficiently.

Experimental results demonstrate improved performance over existing methods.

Abstract

Knowledge discovery in databases aims at finding useful information, which can be deployed for decision making. The problem of high utility itemset mining has specifically garnered huge research focus in the past decade, as it aims to find the patterns from the databases that conform to an objective utility function. Several algorithms exist in literature to mine the high utility items from the databases; however, most of them require large execution time and have high memory consumption. In this paper, we propose a new algorithm, R-Miner, based on a novel data structure, called the residue maps, that stores the utility information of an item directly and is used for the mining process. Several experiments are undertaken to assess the efficacy of the proposed algorithm against the benchmark algorithms. The experimental results indicate that the R-Miner algorithm outperforms the state-of-the-art mining algorithms.