An Efficient Algorithm for Mining Frequent High Utility Occupancy Generators

Discovering frequent high utility occupancy itemsets (FHUOI) in quantitative databases has emerged as a recent trend in data mining. These itemsets are valuable for real-world applications as they capture user preferences and provide a significant utility portion in transactions where they appear. However, the often-enormous size of the set of all FHUOIs, especially in large or dense datasets with low minimum thresholds, presents challenges for analysis. To overcome this limitation, the current study proposes Gen-FHUOIM, a novel algorithm designed to efficiently extract generators of FHUOIs. By utilizing a novel weak lower bound called \(wlbocc\) on utility occupancy and incorporating two new pruning strategies, the algorithm effectively reduces the search space by eliminating FHUOIs that are not generators. Experimental evaluation on various datasets demonstrates that the proposed algorithm outperforms the baseline algorithm for discovering all generators of FHUOIs.