Implementation of Data Mining using Aprirori Algorithm to Improve Sales Patterns of MSME Products

Abstract

Business development is getting faster and full of rivals, especially in the world of MSMEs. MSME entrepreneurs need to develop marketing strategies so that their products can sell a lot so that they can make a profit. One way to assist in marketing strategies can use a priori algorithms so that they can help in sales patterns by using historical consumer transaction data. After testing the results of this system gives good results. This a priori algorithm has several stages in the calculation process starting from the process of calculating support and determining minimum support to evaluate data values that are below minimum support. Then the confident calculation to assist in the formation of association rules so that there are steps to evaluate the value of data that is below the minimum confident. Then there is an elevator to provide information validly and show the transaction process. The results obtained from this a priori algorithm form an association rule "If you buy a cassava balado stick, then buy a balado corn stick" then from this association rule entrepreneurs can be helped in selling products that tend to be bought by many consumers. The last stage is tested from manual calculations ranging from support, confident to elevator with calculations on the web. The test results obtained that manual calculations and calculations on the web have the same value.