IMPLEMENTASI APRIORI PADA PENJUALAN BARANG DENGAN METODE ASOSIASI UNTUK STRATEGI MARKETING

Putra, Josef Cristian Adi (Universitas Bunda Mulia)
Sipayung, Evasaria Magdalena (Universitas Bunda Mulia)

Jurnal Algoritma, Logika dan Komputasi
Vol 7, No 1 (2024): Maret 2024

Technological developments have led to significant changes in various sectors, including business. The way of trading has also gone digital through e-commerce platforms and social media. Business competition is getting tougher with the emergence of many startups. Entrepreneurs must innovate in order to survive the fierce competition. Association analysis is used in Data mining to find rules for combining items. The advantage of this technique lies in the use of efficient algorithms through high-frequency pattern analysis or frequent pattern mining. This algorithm examines candidate itemsets that evolve from the results of frequency itemsets through support-based pruning, to eliminate insignificant itemsets with a Minimum Support value of 1. The Apriori algorithm association method is used to determine item relationships and identify consumer buying patterns, as well as help entrepreneurs increase product sales. This research proves the effectiveness of the Apriori algorithm in managing transaction data and generating valuable information for companies. This research provides input to companies that want to utilize transaction data to improve business effectiveness. The main goal of the Apriori algorithm is to find itemsets that frequently co-occur in the data. The algorithm adopts a bottom-up approach, where smaller itemsets are analyzed first and larger itemsets are built from smaller itemsets. The steps in creating itemsets using the association method include problem identification, transaction data collection, itemset identification, determining the Minimum Support and confidence values, and establishing association rules. This research develops an application that calculates the Apriori algorithm with the associative method through a calculation table and a summary of the calculation results. After testing, the application shows accurate calculation results and can be checked manually. The drawback of this application is that the notification of errors in the data is only displayed one by one.