Perancangan Sistem Informasi E-Commerce Menggunakan Algoritma Naïve Bayes Dan Collaborative Filtering Dalam Peningkatan Akurasi Pencarian Produk

Abstract

This research is motivated by the conventional marketing process of woven fabric at the Lawo Lio MSME in Rateweo Village, Ndori District, Ende Regency, which limits the reach of its products. In addition, sales data management and transaction recording are still carried out manually, making the process less effective and potentially causing recording errors. These conditions indicate the need for a web-based system that can support marketing activities and more systematic sales data management. This study aims to design and develop a web-based e-commerce information system that can help expand marketing reach, facilitate transaction processes, and improve the effectiveness and organization of sales data management at Lawo Lio MSME. The research uses a qualitative method with a descriptive-analytical approach. The system is developed using the Agile method, consisting of planning, implementation, testing, documentation, deployment, and maintenance stages. System analysis and design use a Data Flow Diagram (DFD) to describe the process flow and an Entity Relationship Diagram (ERD) to describe relationships among entities in the database. The system is developed using PHP, JavaScript, and MySQL, while system functionality is tested using the Black Box testing method. The result is a web-based e-commerce information system that provides features for managing customers, products, orders, payments, deliveries, and reports. The system integrates the Naïve Bayes algorithm to support product-search classification and Collaborative Filtering to provide product recommendations based on user preferences. The testing results show that the main system features operate according to the specified requirements. Therefore, the developed system can support the marketing process and the management of woven-fabric sales data more effectively and systematically.