Surprising Ingredient Extraction based on Rarity and Generality

Abstract

Many surprising recipes exist in the user- generated recipe site. The simplest way to find surprising recipe is to use a search function. However, the title of its recipe does not always contain the keyword “surprise”. Thus we cannot find surprising recipes in an easy way. In this paper, we propose a system to extract surprising ingredients as the first step of extracting surprising recipes from the recipe site. We propose RF-IIF (Recipe Frequency-Inverse Ingredient Frequency) based on TF-IDF in our system. RF-IIF calculates a surprising value of the ingredient about the meal based on the generality and the rarity of the ingredient. We extract ingredients whose RF-IIF ranks are in top 20 as surprising ingredients about the meal. Through questionnaires to evaluate the extracted ingredients, we verified the effectiveness of the proposed method which extracts surprising ingredients.