A SWI-Prolog-based expert system for nutrition supplementary recommendation

Abstract

Expert systems have been utilized in recommendation systems for decades. Numerous applications have been successfully developed for many fields. SWI-Prolog is one of the popular auto-reasoning languages for building expert systems. This study presents a SWI-Prolog-based expert system for nutrition supplementary recommendation. The rule base consists of 3 rules and more than 2000 facts. The structure of the rule base is simple, straightforward, and easy-to-implement. Meanwhile, to get better visualization effects, Java has been utilized to build graphical use interfaces (GUIs). These GUIs allow users to input their symptoms. Based on the symptoms users input, the proposed system first diagnoses the shortage of nutrition for the users, and then suggests the foods and vegetables for nutrition supplementary to treat these symptoms via inference procedure based on four different roles: man, woman, pregnant woman, and breastfeeding woman.