Scientists from the Technical University of Denmark (DTU), the University of Copenhagen and Caltech have demonstrated that an algorithm can help consumers match a wine to a person’s taste buds.

To enable this, the team of scientists fed an algorithm with data from people's flavor impressions. This enabled the algorithm to make accurate predictions about what kinds of wine an individual might prefer.

Source: Thoranna Bender

To gather the flavor impression data, researchers held wine tastings wherein 256 participants were instructed to arrange their small cups of different wines on a piece of A3 paper according to which wines they thought tasted the most alike. The scientists, who reasoned that the greater the distance between the cups, the greater the difference in their flavors, digitized the points on the paper by photographing them.

That data was then combined with that culled from hundreds of thousands of wine labels and user reviews obtained through a global wine app to build the algorithm.

The researchers explained: "We can see that when the algorithm combines the data from wine labels and reviews with the data from the wine tastings, it makes more accurate predictions of people's wine preferences than when it only uses the traditional types of data in the form of images and text. So, teaching machines to use human sensory experiences results in better algorithms that benefit the user."

The team added that the algorithm could potentially be transferred to other food and beverages in the future.

The study, “Learning to Taste: A Multimodal Wine Dataset” appears in the journal arXiv.