Almeida, Tracy2026-05-052026-05-052025http://rcca.ndl.gov.in/handle/123456789/548This study presents an enhanced Crow Search Algorithm (CSA) for High-Utility Itemset Mining (HUIM), aimed at improving efficiency in large-scale datasets. HUIM focuses on discovering itemsets based on utility, such as profit or importance, rather than mere frequency. Traditional algorithms struggle with large datasets due to the extensive search space, while metaheuristic approaches like CSA show promise but remain inefficient in handling low-utility itemsets. To address this, we introduce a high-utility item prioritization list to guide the CSA in selecting valuable itemsets, reducing computational complexity. Experimental results, using synthetic datasets, demonstrate that the enhanced CSA significantly reduces execution time as the minimum utility threshold increases. Despite the study’s limitations—such as the use of synthetic data and lack of comparisons with other algorithms—the proposed method showcases potential for practical applications like retail analysis.enArticle