Implementasi Algoritma Apriori untuk Analisis Pola Pembelian Konsumen pada Dataset Market Basket Analysis

Abstract

Penelitian ini bertujuan untuk mengidentifikasi pola pembelian konsumen pada data transaksi penjualan retail menggunakan metode data mining dengan algoritma Apriori. Analisis dilakukan terhadap 20 data transaksi produk kebutuhan harian yang telah dikonversi ke dalam format biner. Proses pengolahan data dilakukan secara manual menggunakan Microsoft Excel serta secara otomatis dengan RapidMiner. Hasil penelitian menunjukkan bahwa produk yang paling sering dibeli oleh konsumen adalah Chocolate (42,1%), Yogurt (42,0%), Butter (42,0%), Ice Cream (41,0%), dan Sugar (40,9%). Selain itu, diperoleh beberapa aturan asosiasi dengan nilai support dan confidence yang tinggi, seperti kombinasi Milk dan Dill → Chocolate (confidence 60%) dan kombinasi Chocolate, Onion, Unicorn → Dill (confidence 60,6%). Informasi ini dapat dimanfaatkan untuk menyusun strategi penjualan yang lebih efektif, seperti pembuatan paket bundling produk, penempatan produk secara strategis di rak toko, serta pengembangan sistem rekomendasi berbasis pola pembelian konsumen.   Kata Kunci: Data mining, Apriori, Market Basket Analysis, Association Rules, RapidMiner.   This study aims to identify consumer purchasing patterns in retail transaction data using data mining techniques with the Apriori algorithm. The analysis was conducted on 20 daily product transactions that were converted into binary format. Data processing was carried out both manually using Microsoft Excel and automatically with RapidMiner. The results indicate that the most frequently purchased products by consumers include Chocolate (42.1%), Yogurt (42.0%), Butter (42.0%), Ice Cream (41.0%), and Sugar (40.9%). Furthermore, several association rules with high support and confidence values were discovered, such as the combination of Milk and Dill → Chocolate (confidence 60%) and Chocolate, Onion, Unicorn → Dill (confidence 60.6%). These insights can be utilized to design more effective sales strategies, such as bundled product promotions, optimized shelf arrangements, and the development of recommendation systems based on consumer purchasing behavior.  Keywords: Data Mining, Apriori, Market Basket Analysis, Association Rules, RapidMiner