An Architecture for Food Product Recommendation Focusing on Nutrients and Price

Abstract

Part of the world’s population is nutrient deficient, a phenomenon known as hidden hunger. Poor eating conditions cause this deficiency, leading to illnesses and recovery difficulties. Malnourished patients are more easily affected by Covid-19 and have a difficult recovery after the illness. An effective food choice has the price and nutritional value of food products as the most relevant factors, with the price being the most relevant, considering the context of countries such as Brazil. Thus, having identified a scenario in which the access and food price mainly cause malnutrition. This work proposes an architecture, called Nutri’n Price, to recommend high nutritional foods with low costs. The architecture encompasses a network of ontologies, inference algorithms, information retrieval and collaborative filtering techniques to recommend the best foods according to nutrient choice, price, and user contextual information. A prototype of a mobile application was developed to evaluate the feasibility of the proposed architecture.