Belgian beer study acquires taste for machine learning

Researchers reckon results could improve recipe development for food and beverages

Joining the list of things that probably don't need improving by machine learning but people are going to try anyway is Belgian beer.

The ale family has long been a favorite of connoisseurs worldwide, yet one group of scientists decided it could be brewed better with the assistance of machine learning.

In a study led by Michiel Schreurs, a doctoral student at Vlaams Instituut voor Biotechnologie (VIB) in Flanders, the researchers wanted to help develop new alcoholic and non-alcoholic beer flavors with higher rates of consumer appreciation.

Understanding the relationship between beer chemistry and its taste can be a tricky task. Much of the work is trial and error and relies on extensive consumer testing.

The Belgian group documented 200 chemical properties from 250 commercial Belgian beers across 22 styles, such as Blond and Tripel beers. Using online beer review database RateBeer, they then linked 180,000 sensory descriptions to known chemical compositions. They also used profiling data from a trained tasting panel of 16 people.

Drawing on the dataset, the team trained machine learning models to correlate and predict flavor in terms of how the drinker's appreciation correlates with the chemical profile.

According to a paper published in Nature Communications today, the machine learning model achieved higher overall appreciation among trained panelists in blind tastings.

"The flavor of beer is a complex mix of aroma compounds," Schreurs said. "It is impossible to predict how good a beer is by measuring one or a few compounds. We really need the power of computers."

The authors suggest this tool could help improve quality control and recipe development of many types of food and beverages to help meet consumer demands more efficiently, but the study was limited to Belgian beers.

Corresponding author Kevin Verstrepen, professor at Belgian university KU Leuven, said one of the biggest goals is now to help make better alcohol-free beer. "With the use of our model, we already succeeded in creating a cocktail of natural aroma compounds that mimic the taste and smell of alcohol, without the risk of a hangover," he said in a statement.

Schreurs admitted that the team did celebrate finishing the paper with the alcohol-containing variety. With some Belgian beers touching 15 percent alcohol by volume, he didn't say how bad the hangover was. ®

More about


Send us news

Other stories you might like