Fig. 1
From: The recipe similarity network: a new algorithm to extract relevant information from cookbooks
The recipe network assembling. (A) The process starts from the recipe sets. A recipe is a set of individual ingredients from which we compute the intersection set with another recipe. (B) The intersection set among recipes represents the common ingredients. If there are no common ingredients between two recipes, the intersection is an empty set. Each non-empty intersection corresponds to a network link connecting the two recipes sharing common ingredients in the intersection graph; (C) the intersection graph of recipes in which each link represents an intersection set of ingredients; (D) the trivial weighted network of recipes extracted from the intersection graph in panel (C); here, link weight indicates the number of common ingredients between recipes, i.e., the cardinality of the intersection; (E) the recipe similarity network (RSN) in which we compute the link weight using Eq. 3; the RSN is an undirected and weighted network where the link weight indicates the normalised average measure of ingredient similarities between recipes.
