Sistem Informasi Penjualan Barang Berbasis Web Menggunakan Algoritma Naive Bayes ( Studi Kasus di Toko Yakusa-Ende Tengah)

Abstract

The Yakusa Store, located in Central Ende, is a retail business that still manages sales data manually using notebooks, which often results in recording errors, difficulties in stock management, delays in reporting, and an inability to analyze sales patterns of items favored by customers. This study aims to develop a web-based sales information system integrated with a predictive algorithm to address these issues. This study applies the waterfall method with the stages of needs analysis, system design, implementation, verification, and maintenance. Data collection was conducted through literature review, observation, and interviews with the store owner. The system was designed using PHP, MySQL, and XAMPP. The Naive Bayes algorithm was applied to classify sales levels into three categories: best-selling (high), normal (medium), and less popular (low) based on historical transaction data. The system was tested using the black box testing method, and the Naive Bayes algorithm achieved an accuracy of 83.3%, indicating that the system is effective for predicting the sales level of goods at Toko Yakusa.