Um Sistema de Recomendação para Nutrição Oncológica baseado em Decisão Multicritério

Abstract

Oncological patients face metabolic, immunological, and biochemical challenges that often lead to malnutrition, compromising both treatment effectiveness and quality of life. In this context, a personalized and multidisciplinary approach, such as the use of recommendation systems, emerges as a valuable tool to assist nutrition professionals in developing dietary plans tailored to each patient’s specific needs. This study applies a multi-criteria recommendation system for oncological patients, utilizing nutritional data from foods and synthetic data generation to create personalized profiles. The computational models developed, using the TOPSIS and NSGA-II approaches, successfully generated dietary recommendations that meet individual nutritional goals while adhering to all necessary restrictions. Additionally, a method for synthetic data generation was implemented, enabling the creation of personalized oncological patient profiles. Complementing the work of nutrition professionals, the proposed approaches can contribute to dietary recommendations for oncological patients.