Analisis Pola Penjualan Produk Diet Menggunakan Metode Apriori

Abstract

Healthynoona Store is a company engaged in the field of diet products in Surabaya. Its daily operational activities have increased the number of transactions, while the existing transactions have not been neatly organized. Currently, the problem in the Healthynoona store is related to the availability of very large sales data, which is not used optimally because there is no decision support system to design a business strategy to increase sales. Consequently, data mining is needed to reduce the number of risks that are detrimental to the company. This system applies the Apriori algorithm method by determining sales patterns at the Healthynoona Store and can help find out which products consumers often buy simultaneously. The test results based on transaction data for 3 months in April–June 2020, with a minimum support of 15%–65% and a minimum confidence of 50%–100%, produced 11 tests. Meanwhile, a minimum support of 25% and a minimum confidence of 45% of transaction data for 1-6 months in April–September 2020 produced 6 tests. Thus, the system has been running well. In conclusion, the greater the minimum support and minimum confidence used, the less or even no association rule results would be found. Investigating the number of transactions processed would affect the association rules, and their number would vary.