<![CDATA[The roast level of coffee has a significant effect on its taste, aroma, and viscosity. However, it is often determined subjectively by roasters based on personal experience, especially among small and medium-sized coffee businesses. This study aims to design a web-based coffee roast level recommendation system using a Rule-Based System approach. The system was designed using the Waterfall method and implemented in PHP programming language with a MySQL database. The main attributes used in the system include coffee type, serving method, post-harvest process, and flavor profile. The roaster's knowledge is represented in 20 IF–THEN rules and executed using the Forward Chaining technique. The result of this research is a recommendation system that is able to provide consistent and objective advice on coffee roast levels, such as Light Roast, Medium Roast, and Dark Roast. This system is expected to help roasters or SME players improve the quality of coffee serving in a more standardized and efficient manner.]]>
Copyrights © 2025
