Preserving Privacy in Association Rule Mining Using Metaheuristic-Based Algorithms: A Systematic Literature Review

The current state of Association Rule Mining (ARM) technology is heading towards a critical yet profitable direction. The ARM process uncovers numerous association rules, determining correlations between itemsets, forming building blocks that have led to revolutionary scientific discoveries. However, a high level of privacy is vital for protecting sensitive rules, raising privacy concerns. Researchers have recently highlighted challenges in the Privacy-Preserving Association Rule Mining (PPARM) field. Many studies have proposed workarounds for the PPARM dilemma by using metaheuristics. This paper conducts a systematic literature review on metaheuristic-based algorithms addressing PPARM challenges. It explores existing studies, providing insights into diverse metaheuristic approaches tackling PPARM problems. A detailed taxonomy is presented, offering a structured classification of metaheuristic-based algorithms specific to PPARM. This classification facilitates a nuanced understanding of the field by categorizing these algorithms into metaphor-based and non-metaphor-based groups, with a discussion of the nature of the representation schemes for each category identified in the survey. The review extends its analysis to encompass the latest applied approaches, highlighting the diversification of existing metaheuristic algorithms in the PPARM context. Moreover, common datasets and evaluation metrics identified from selected studies are documented to provide a deeper understanding of the methodological choices made by researchers in this domain. Finally, a discussion of existing challenges and potential future directions is presented. This review serves as a helpful guide that outlines previous research and presents potential future opportunities for metaheuristic-based algorithms in the context of PPARM.