<![CDATA[Order cancellation is one of the operational problems in food delivery applications because it can affect service efficiency, restaurant preparation processes, delivery partner productivity, and customer satisfaction. This study aims to analyze order cancellation patterns in food delivery applications using the Apriori algorithm. The research used a quantitative data mining approach with a simulated food delivery transaction dataset consisting of 1,000 records. The dataset included transaction attributes such as order time, restaurant category, payment method, promotion status, delivery distance, delivery fee, estimated waiting time, and order status. The data were processed through cleaning, transformation into categorical itemsets, frequent itemset generation, and association rule formation. The Apriori algorithm was used to identify relationships between transaction attributes and cancelled orders based on support, confidence, and lift values. The results showed that cancellation patterns were strongly associated with long estimated waiting time, far delivery distance, and high delivery fee. The strongest rule was the combination of long waiting time and far delivery distance leading to cancelled orders, with support of 0.120, confidence of 0.750, and lift of 2.50. These findings indicate that the Apriori algorithm can generate interpretable association rules that are useful for identifying high-risk order conditions. The results can support food delivery platform managers in improving service efficiency, reducing cancellation risk, and making data-driven operational decisions.]]>
Copyrights © 2025
