Application of the Apriori Algorithm to Purchase Patterns

Abstract

The purpose of this research is to produce an Apriori Algorithm application system to increase sales turnover at Viona stores. The problem faced by the Viona store is that the Viona store has decreased turnover in the midst of business competition because it has not been able to optimally analyze the products that are often purchased and the combination of purchases by consumers so that sales seem monotonous and do not have a business strategy to attract customers. sales that can attract consumers. One way is to make sales with sales packages at lower prices. It must have a good pattern and analysis to be able to combine products into a sales package. However, with the limited ability of Information Technology, in this study a sales application was built that applies the a priori algorithm. This a priori algorithm is very effective in finding the relationship pattern of one or more itemsets in a large data set so that it is effective in calculating a sales transaction data and finding patterns of combinations of consumer habits and being able to quickly create product sales packages. increase sales turnover. The results of the process of applying the a priori algorithm to sales data at the Viona Store through the RapidMiner application are the same as the results applied to the system built and using sales transaction data for the month of May 2022 using a minimum support of 30% and minimum confidence of 30%. So from this study, information was obtained that the items that were often purchased together during this May period were lighters and cigarettes with 100% Confidence. And for the month of June Viona Stores can recommend packages in their store by looking at the results of a combination of 3 items, which are later expected to increase sales turnover at Viona Stores.