Gaby
German researchers have developed two machine learning algorithms that can easily determine the country of origin of various types of whiskey and analyze the aromas of these beverages. As is well known, the aroma of whiskey is characterized by a complex blend of fragrant compounds, making its evaluation quite challenging.
The study’s results showed that the artificial intelligence system even outperformed human experts in terms of accuracy.
What Did the Researchers Report?
Typically, involving human sommeliers in whiskey evaluations requires significant time and financial investment. Moreover, consensus among experts is a rare occurrence.
To help the whiskey market avoid both material and moral costs, a team of scientists from the Fraunhofer Institute for Process Engineering and Packaging (Germany) developed an AI system for expert whiskey evaluation, named OWSum.
During testing, it was tasked with determining the molecular composition of seven types of American whiskey and nine types of Scotch whiskey.
The artificial intelligence analyzed samples of the noble drink and identified the country of origin for each type, along with its five strongest flavor notes.
Subsequently, the authors of the study compared the results from the AI with a control group of 11 experts, as reported by the Daily Mail. It turned out that OWSum can accurately determine whether a whiskey is American or Scotch with over 90 percent accuracy.
The AI identified caramel flavor as the most characteristic note of American whiskey, while apple and phenolic notes (often described as smoky or medicinal) were recognized as the most distinctive notes of Scotch whiskey.
Both algorithms were able to pinpoint the five strongest notes of a specific whiskey more accurately and confidently than any expert in the control group.
The results of the study were published in the journal Communications Chemistry.
