<![CDATA[Digital transformation opens strategic opportunities for Micro, Small, and Medium Enterprises (MSMEs) to expand market reach through electronic commerce platforms. Plaza Banten is a digital marketplace that facilitates the promotion and sale of MSME products in Banten Province, Indonesia. Although transactions on the platform continue to increase, the utilization of transaction data for strategic decision-making has not yet been optimized. This study aims to identify consumer purchasing patterns on Plaza Banten through data mining by discovering product associations frequently purchased together and translating them into recommendations for promotions, product placement, and inventory planning. Sales transaction data were collected for a specific period and preprocessed through cleaning, transformation, and relevant attribute selection. The Apriori algorithm was applied in two scenarios: overall analysis and time-based transaction segmentation. Using a minimum support of 0.1% and minimum confidence of 60%, the analysis generated 8,117 association rules. The strongest rule achieved support = 0.348 and confidence = 98.9% (Nasi Box → Snack Box), while several segments reached confidence up to 100%. The highest lift value was 194.75 in the 06:00–09:00 segment, indicating highly specific co-purchase dependencies at certain times. These quantitative results reveal stable bundle patterns and time-dependent demand variations, supporting actionable strategies such as standardizing menu bundles, optimizing cross-selling offers, and prioritizing stock for high-correlation items. The resulting rules are interpreted and visualized to support Plaza Banten administrators and MSME partners in implementing data-driven decisions and strengthening the digital economy ecosystem in Banten Province.]]>
Copyrights © 2026
