Market Basket Analysis Using Apriori and Eclat Algorithm in an E-Commerce Company

Abstract

E-commerce companies are facing significant challenges due to escalating competition, the homogenization of products and services, rapid shifts in customer demands, and the burgeoning volume of customer transactions. Consequently, these companies grapple with complex problems such as customer segmentation, customer churn analysis, market basket analysis, and the design of product recommendation systems. Leveraging data mining and machine learning algorithms offers substantial opportunities to effectively address these issues. For e-commerce enterprises, analysing customers’ purchasing behaviours and crafting personalized product recommendations not only enhances customer loyalty but also encourages impulse purchases. This approach also contributes to a more user-friendly platform, thereby increasing customer satisfaction. The present study aims to conduct a market basket analysis utilizing real-world data from an e-commerce company. Association rule mining algorithms, specifically Apriori and Eclat, are employed for this analysis. Through these methods, correlations and patterns between product categories are uncovered. Moreover, each algorithm is analyzed by comparing its execution duration and the quality of its results. These specified techniques, including Apriori, an intelligent join-based algorithm, and Eclat, a sophisticated tree-based algorithm, demonstrate remarkable intelligence by efficiently identifying and analyzing patterns within complex datasets. Their innovative methodologies enable them to dynamically adapt to data structures and extract frequent itemset with high performance, as evidenced by their outstanding results in current literature.