Sistema de recomendação de receitas alimentares utilizando filtragem baseada em conteúdo

Abstract

In recent years, due to the lack of time, the difficulty to maintain a healthy diet is increasing more and more. Because of this, coupled with the fact that the world is experiencing a pandemic moment, psychological problems such as anxiety and depression are becoming more common. In addition, the amount of food recipes available on the internet has been growing at an accelerated pace. Taking this into account, this work propose a study to develop a prototype of a recipe recommendation system that considers the ingredients and their quantities. In this way, the user will be able to find recipes faster and, with this, will be able to develop the act of cooking as a hobby. To achieve this, the prototype uses content-based filtering to generate the recommendations. The experimentation takes place with the system generating five recommendations based on the euclidian distance from the recipe choosen. The experiments have shown that the prototype generated satisfactory recipes for the users.