<![CDATA[An increase in sales transaction data every day will cause a very large amount of product sales transaction data to be stored. In most cases, data relating to very large sales transactions are only stored and used for archival purposes, not exploited adequately. In retail marketing, association data mining is used to investigate product purchasing patterns. However, if the company does not know the customer's purchasing patterns, it can have impacts such as inappropriate marketing strategies, decreased customer retention, lost business opportunities, lack of personalization, tough competition, stock/production inefficiencies, loss of customer trust. Knowing consumer purchasing patterns, the company can develop sales strategies and make the right decisions. In this study using Market Basket Analysis using the Apriori Algorithm and FP-Growth to determine consumer buying patterns. The results of this study resulted in two itemset combinations. The first combination is that if the buyer buys yogurt and sausage, the buyer also buys whole milk. The resulting support value is 0.00147 (0.0147%), the confidence value is 0.255814 (25.58%) and the lift value is 1.61986. the second combination, namely if the buyer buys sausage (sausages) and rolls/buns (bread rolls), then also buys whole milk (milk), this combination produces a support value of 0.001136 (0.0113%), a confidence of 0.2125 (21.25%) and a lift of 1.34559 . In addition to the combination of the itemset produced in this study, it also measures computational speed in processing Groceries data for Market Basket Analysis. The computational speed produced by the Apriori Algorithm is 3.1765 seconds, while the FP-Growth algorithm is 0.15892 seconds. The difference in computational speed between the Apriori Algorithm and FP-Growth is 3.0176 seconds.]]>
Copyrights © 2023
