Intelligent Generation of Flavor Preserving Alternative Recipes for Individuals with Dietary Restrictions

Abstract

Many diseases lead to treatments which bring forth certain restrictions in diet and quantity of consumption of the patients. Our target has been to work around these restrictions to generate a novel approach for creating a balance in taste and flavor while maintaining the nutritional properties within the given upper bound. The resulting product is not only a flavor-wise approximate supplement of the item to be replaced but is also created from ingredients which are derived from the list of food items permitted for consumption by a particular patient suffering from a certain disease. There is a two layer selection procedure which chooses the substitutes from the given list based on the flavor profile of the item whose supplement is to be found, with the cut-off being designed on the basis of the calories, carbohydrates, sugars and fats permitted for the patient being considered to consume. Our results have shown to predict near-identical flavor profiles with significantly lower values of calories, fats, sugars or carbohydrates as per the requirement of the patient. The output has been validated by experienced food technologist.