Sure chefs and gourmands are the food experts that can experiment and create new and interesting combinations, but it seems that they may well be assisted in this task by Artificial Intelligence. A newly published paper by scientists at the University of Korea (“KitcheNette: Predicting and Recommending Food Ingredient Pairings using Siamese Neural Networks“), describes a machine learning system that predicts food and drink pairings and rates them by score.
"Many chefs, gourmets, and food-related researchers have focused on studying food pairing for decades,” wrote the paper’s authors. “Since food pairings are made based on the experiences of experts, food pairing itself is subjective and difficult to quantify. In this work, we introduce KitchenNette, which … predicts the scores of unknown pairings consisting of food ingredients that have infrequently or never been used in recipes.”
The system is based so-called Siamese networks, or two identical twin machine learning models that take two differing datasets as inputs. To prime the models scientists provided one model with 356,451 known ingredient pairings from recipes (text and images). The other model was able to score or rate the pairings based on the first set.
To test the models, scientists took three different sparkling wines – Champagne, Prosecco and sparkling wine and then calculated the score based on pairing with a certain ingredient. As expected parings like Champagne and orange twist scored consistently high while odd pairing such as Prosecco and onion scored low. The system also successfully predicted high scoring food pairings that fall into the classic gastronomic class.
The next phase for the model is to feed it chemical information about food and more detailed descriptions from encyclopaedias as well as using more ‘novel’ and ‘authentic’ recipes to help the model recommend more ‘versatile’ food pairings.
AI in food and recipe creation is ramping up and, while we won't say that the robots are coming for chefs jobs just yet, it does look like tech, algorythms and AI will be a large part of the future for chefs.