MFG-HUI: An Efficient Algorithm for Mining Frequent Generators of High Utility Itemsets

Abstract

The discovery of frequent generators of high utility itemsets (FGHUIs) holds great importance as they provide concise representations of frequent high utility itemsets (FHUIs). FGHUIs are crucial for generating nonredundant high utility association rules, which are highly valuable for decision-makers. However, mining FGHUIs poses challenges in terms of scalability, memory usage, and runtime, especially when dealing with dense and large datasets. To overcome these challenges, this paper proposes an efficient approach for mining FGHUIs using a novel lower bound called $lbu$ on the utility. The approach includes effective pruning strategies that eliminate non-generator high utility branches early in the prefix search tree based on $lbu$, resulting in faster execution and reduced memory usage. Furthermore, the paper introduces a novel algorithm, MFG-HUI, which efficiently discovers FGHUIs. Experimental results demonstrate that the proposed algorithm outperforms state-of-the-art approaches in terms of efficiency and effectiveness.