Nutritional Menu Planning: A Hybrid Approach and Preliminary Tests
Oscar Chávez-Bosquez, Jerusa Marchi, Pilar Pozos-Parra · Research in Computing Science · 2014
Menu planning is a process appearing to be straightforward but many complexities arise when it is tried to be solved by computer means. Actually, although there is evidence of previous work since 50 years ago, at present there is no wide know tool which can solve this task in an automated manner. Also, not all proposals deal with full recipes along with considering the user food preferences. In this paper we pro- pose a system architecture based on hybrid optimization: a rst module based on mathematical programming, a well known robust approach to this problem; and a second module based on belief merging, a lesser known framework aimed to combine the nutrition scientist advices and policies along with the user food desires. The association of numerical and symbolic approaches will allow us to generate of a more agreeable menu. In order to illustrate our proposal, we present a motivating example detailing the main aspects of the system.