There has been a significant advancement in artificial intelligence (AI) related to wine tasting. Researchers at the University of Copenhagen’s Pioneer Centre for AI, along with collaborators from the Technical University of Denmark (DTU) and Caltech, have successfully trained an algorithm to predict individual wine preferences based on human taste perceptions.
The study explains how this was achieved through wine-tasting events where participants ranked wines based on their taste similarities. The data from these events was combined with extensive wine label and user review information provided by the wine app Vivino. This includes 897,000 images of wine labels and 824,000 reviews covering over 350,000 unique vintages.
The key innovation here is the incorporation of human sensory experiences into AI, allowing the algorithm to make more accurate predictions about wine preferences. The study opens new avenues for understanding the subtle nuances of flavour. The algorithm can be used in areas such as marketing, product development and customer satisfaction.
