Penerapan Algoritma FP-Growth untuk Analisis Pola Pemesanan Menu pada Transaksi Restoran

Advances in information technology enable restaurants to utilize transaction data to support business decision-making. However, many restaurants have not yet optimally processed transaction data, resulting in a lack of understanding of customer ordering patterns. This study aims to analyze menu ordering patterns in restaurant transaction data using the FP-Growth algorithm. The CRISP-DM method used encompasses the stages of Business Understanding, Data Understanding, Data Preparation, Modeling, Evaluation, and Deployment. Transaction data was processed and analyzed using the RapidMiner application to identify relationships between frequently ordered menu items. The results revealed several association patterns, such as the relationship between Cheeseburger and Hot Dog with a confidence value of 0.400 and Hot Dog and Veggie Burger with a confidence value of 0.400. These patterns indicate a tendency for customers to choose certain menu combinations in a single transaction. The resulting information can be used by restaurants to develop promotional strategies, create menu packages, conduct cross-selling, and manage raw material inventory more effectively. Thus, the FP-Growth algorithm has proven effective in identifying customer ordering patterns and supporting restaurant business decision-making.