Sign In

Communications of the ACM

ACM TechNews

Deep Learning Algorithm Rewrites Traditional Recipes For New Regions, Ingredients

View as: Print Mobile App Share:
Algorithmic gastronomy is here.

A new machine learning algorithm developed by a team of French, American, and Japanese researchers offers an automated way to adapt cuisines according to the culinary traditions of some other place.

Credit: Shutterstock

A team of French, American, and Japanese researchers says it has developed a new machine-learning algorithm that can take a given recipe and transform it into an alternative dietary style.

The researchers say the system initially analyzes a large number of recipes and uses them to train a neural network as to what recipe features represent the culinary style of a given country or region. The model can then be used to take new lists of ingredients and make predictions as to the countries' cuisines to which that list corresponds.

The next step involves taking the recipe's style mixture and visualizing it as a Newton diagram, which displays mixtures within data as two-dimensional coordinates.

The third system element clusters ingredients together using a word2vec model, which is a way of quantifying associations between words and results in an interpretation of how similar ingredients are to one another and which ingredients can be substituted for other ingredients.

From Motherboard
View Full Article


Abstracts Copyright © 2017 Information Inc., Bethesda, Maryland, USA


No entries found

Sign In for Full Access
» Forgot Password? » Create an ACM Web Account