<![CDATA[Micro, Small, and Medium Enterprises (MSMEs) are vital to Indonesia's economy but often face challenges in inventory control and understanding consumer behavior. This study aims to compare the performance of the Apriori and FP-Growth algorithms in identifying consumer purchasing patterns from 7,778 transaction records at UD. Kurnia, a building material store, between August 2023 and July 2024. Unlike previous research that relied only on support and confidence metrics, this study applies the lift metric, which measures the strength of item associations, to minimize misleading rules. The algorithms were tested under 15 combinations of minimum support and lift threshold values. Results show that both algorithms generate the same association rules, but Apriori is significantly faster. At a minimum support of 0.0005 and a lift threshold of 1.5, Apriori completes processing in 3.23 seconds, while FP-Growth takes 21.81 seconds. With these findings, store owners can make more precise inventory decisions and implement data-driven cross-selling strategies, such as offering semen gresik when colt pasir is purchased.]]>
Copyrights © 2026
