Menu Package Recommendation Menu Package Recommendation Using Combination of K-Means and FP-Growth Algorithms at Bakery Stores

N P Dharshinni (Universitas Prima Indonesia)
Elvana Bangun (Universitas Prima Indonesia)
Sarah Karunia (Universitas Prima Indonesia)
Ruth Damayanti (Universitas Prima Indonesia)
Gabriel Rophe (Universitas Prima Indonesia)
Roy Pandapotan (Universitas Prima Indonesia)

Jurnal Mantik
Vol. 4 No. 3 (2020): November: Manajemen, Teknologi Informatika dan Komunikasi (Mantik)

Bakery shop is a shop that sells variants menu like bread, cakes, and drinks. The main problem with this store's sales is still not knowing which product items are best sellers and the shop still markets a lot of non-selling menus, causing the shop to lose money. So it takes the right strategy to increase the sales of bakery shop menus by making a menu package recommendations from the menus most frequently purchased by customers. The k-means algorithm performs grouping on menus to get menu packages. Furthermore, the fp-growth algorithm looks for linkages between frequently purchased menus to get menu package recommendations. The results of the research that the dominant items often purchased in cluster0 packages are hotdogs, pancakes, milk, garlic breadsticks with a confidence value of 92%, cluster1 packages are garlic breadsticks, hotdogs, chicken sand, pancakes with a confidence value of 92% and the last cluster2 packages are garlic breadstick, pastry, milk with a confidence value of 79%.