Cynthia Marling, Grace Petot, and Leon Sterling
The task of designing nutritious, yet appetizing, menus is one at which human experts consistently outperform computer systems. Tailoring a menu to the needs of an individual requires satisfaction of multiple numeric nutrition constraints plus personal preference goals and aesthetic criteria. We have combined case-based reasoning (CBR) and rule-based reasoning (RBR) in a hybrid system which designs a daily menu for an individual, in accordance with nutrition guidelines, personal preferences, and aesthetic criteria. The hybrid system incorporates the strengths of independent CBR and RBR systems built to perform the same task. CBR is used to satisfy multiple constraints, while RBR allows the introduction of new foods into menus and the performance of "what if" analysis needed for creative design. The CBR/RBR hybrid outperforms either single strategy system. It provides a new framework for planning special purpose therapeutic menus, such as those for diabetics and cardiac patients. We believe that the hybrid approach would extend to other design domains in which both physical constraints and aesthetic considerations are important.