Developing an expert system for diet recommendation

Abstract

Menu construction is an important task for institutions that need to plan menus within certain constraints. There is also a personal need for professional menu construction by clients or patients who should eat according to a planned diet. For menu construction and dietary analysis, there are several approaches (e.g., linear programming, genetic algorithms, rule-based expert systems, etc.) and commercial IT systems. In this paper, we propose a case-based approach for diet recommendation. Based on this approach, we are going to construct an expert system which is intended to be employed in a health record management system. Our approach is based on ripple down rules (RDR), however, a special representation is also needed for patient attributes and rule actions.