<![CDATA[The development of digital transactions in the retail industry has generated large volumes of data containing valuable information regarding consumer purchasing patterns. This study aims to identify association rules among products using the association rule mining method with the Apriori algorithm to support cross-selling strategies and product arrangement optimization. The dataset consists of 3,898 transactions with 167 unique products and exhibits sparse market basket characteristics with a density of 5.95%, indicating that most product combinations rarely occur together. The research employs the CRISP-DM framework, which includes data understanding, data preparation, modeling, and evaluation stages. In the modeling stage, the Apriori algorithm was applied with a minimum support threshold of 0.01 and a minimum confidence threshold of 0.40, resulting in 3,016 frequent itemsets and 3,398 association rules. After filtering using the criteria of lift > 1.0 and confidence >= 0.40, a total of 2,228 rules met the quality standards. Validation using the Chi-Square test showed that 74% of the rules were statistically significant at a 95% confidence level (p-value < 0.05). One of the best-performing rules indicates that the purchase of Other Vegetables, Rolls/Buns, and Yogurt has a strong relationship with the purchase of Whole Milk, with a support value of 0.0344, confidence of 65.69%, and lift of 1.434. This study contributes through the implementation of multi-metric evaluation and statistical validation to improve the reliability of association rules. The findings can be utilized by the retail industry for shelf arrangement strategies, bundling promotions, and data-driven cross-selling optimization.]]>
