Analysis of Consumer Purchasing Behavior in Groceries Through Apriori-Based Market Basket Analysis

Abstract

The rapid growth of digital transaction systems in grocery retail has resulted in large volumes of transactional data that can be leveraged to understand consumer purchasing behavior. However, extracting meaningful and actionable insights from such data remains a challenge due to the complexity and scale of transaction records. This study addresses this problem by applying Apriori-based market basket analysis to identify association rules that represent purchasing patterns among grocery products. The motivation of this research lies in the increasing need for interpretable and data-driven decision support tools that can assist retailers in optimizing product placement, bundling, and promotional strategies. The proposed approach systematically processes point-of-sale transaction data through preprocessing, frequent itemset generation, and association rule extraction. The contribution of this study is twofold: first, it provides an empirical analysis of consumer purchasing behavior using real grocery transaction data; second, it demonstrates the effectiveness of Apriori-based market basket analysis combined with multiple interestingness metrics to ensure rule relevance and interpretability. The evaluation is conducted using standard metrics, including support, confidence, lift, and complementary measures, to assess both statistical strength and business relevance of the extracted rules. The results show that several product categories exhibit strong associative relationships, offering valuable insights for retail strategy formulation. Future work may extend this study by incorporating temporal analysis, customer segmentation, or comparative evaluation with advanced association rule mining algorithms to further enhance analytical depth and practical applicability.