Comparative Analysis of Apriori, FP-Growth, and ECLAT Algorithms for Effective Grocery Store Management and Product Recommendations

In the dynamic landscape of grocery retail, administrators face the challenge of efficiently managing inventory and making informed decisions to enhance sales and customer satisfaction. Traditional methods of inventory management and product recommendation may not fully leverage the wealth of transactional data available. This study aims to tackle these obstacles through the introduction of a Grocery Recommendation System that utilizes Market Basket Analysis (MBA). The goal of this system is to successfully recommend products and anticipate client preferences by utilizing transactional data taken from the custom e-commerce website. The website is made to record a variety of transactional data when a customer makes a purchase. By using sophisticated algorithms like Apriori, FP-Growth (Frequent Pattern Growth), and ECLAT (Equivalence Class Clustering and Bottom-Up Lattice Traversal), the system detects recurring itemsets and association rules that emphasize the significant connections across products. This study evaluates the performance of all three algorithms, focusing on metrics like time complexity, support, confidence, and lift. FP-Growth demonstrates superior time efficiency, particularly with large datasets, due to its memory-efficient frequent itemset generation. Apriori's straightforward solution, meanwhile, might be appropriate for smaller datasets that need to be interpretable. ECLAT’s vertical transaction method offers advantages in dense data scenarios. Selection among Apriori, FP-Growth, or ECLAT should align with specific dataset characteristics and business goals. These algorithms provide actionable insights crucial for inventory management and customer engagement strategies. Administrators are then able to increase inventory turnover, adjust stocking techniques , and eventually increase sales.