Recipe-transition graph based on asymmetric entropy difference

Abstract

There are many recipes that are regularly shared on the Internet, and they comprise features such as ingredients, cooking procedures, and cooking time. However, it is sometimes difficult to understand the recipes exactly because they are not always written quantitatively and objectively. In order to preserve the accuracy of the recipes, an objective approach is required. In this paper, we define the quantitative characteristics of recipes and determine the relation between them. We use the ingredients of recipes to define their characteristics. We define each of the ingredients based on the statistical entropy measure. In addition, we develop a recipe-transition graph, and we define the asymmetric entropy difference by considering the difference in the ingredients of two recipes as the transition effort. We then show the utilization of this recipe-transition graph. The experimental results show that the most transitive recipes and the recipe transition path can be found with the recipe-transition graph.