A Study of Diversity Analysis Model Based on Embeddings for Cooking Recipes

Recently, a large number of cooking recipes have been posted and shared on the Internet. Various machine learning techniques have been proposed to analyze those recipes. Those include a method to discover alternative ingredients by obtaining distributed representations from cooking procedures and ingredient names, or a method to extract basic procedures from common features in cooking procedures. Such methods utilize the constructed semantic space to calculate the distances among cooking procedures and ingredients for recipes, and demonstrate effectiveness by evaluating similarity of recipes. Using a similar semantic space, we can analyze not only the similarities among recipes but also their diversity. Even for the same dish name, there could be a variety of recipes, depending on the contributor. The diversity of recipes varies from dish to dish. By taking this diversity into account, it is possible to perform various analyses such as extracting recipes that are suitable for each user. In this study, we propose a method to analyze the diversity of recipes using distributed representation. In addition, we apply the proposed method to the posted data on an actual recipe site and show its usefulness.