<![CDATA[Roma Bangunan is one of the mainstay building stores of the Karawang community, offering a wide range of building material products. However, this store faces the challenge of unstable stock inventory which hampers operations and service to customers. Inconsistent stock can lead to service delays, lost sales, and reduced customer satisfaction. Therefore, the author uses Rapidminer with the application of a priori algorithm to find sales combination patterns in sales transaction data, so that this combination pattern will later become information on what goods must have more and less stock of goods, thus minimizing the occurrence of overstocking. In this study, 3 association rules/combination patterns were produced that met the minimum support of 10% and confidence of 50%, namely for the highest association rule "If consumers buy sandpaper, then buy White Brush" with a support value of 12%, confidence of 69% and an elevator ratio of 3.11. By implementing Rapid Miner and a priori algorithms, Roma Bangunan Stores was able to leverage existing data to make better decisions, optimize inventory management, and ultimately improve customer satisfaction and overall business performance.]]>
