This tool is expected to help reduce fraud and exaggeration of wine quality in the market, The Guardian reported today, December 4.
The researchers used ultra-sensitive technology to differentiate wines based on even minute differences in the concentrations of compounds, allowing them to trace wines not only to a specific wine-growing region, but also to the winemaking region.
New technology is expected to help reduce the situation of "false advertising" in the wine market.
Professor Alexandre Pouget at the University of Geneva in Switzerland said there were many cases of wine fraud, with people making up some nonsense, printing labels and selling them for thousands of dollars. “For the first time we have shown that we have the sensitivity with chemical techniques to tell the difference,” he said.
To train the AI program, the scientists used a method called gas chromatography. This method was used to analyze 80 wines harvested over 12 years, from seven different estates in the Bordeaux region of France. This technique is commonly used in laboratories to separate and identify the compounds that make up a mixture.
Instead of trying to find individual compounds to distinguish one wine from another, the algorithm relies on all the chemicals detected in the wine to find the most reliable signature for each type.
The program then returns the results in a tabular format to the user, where wines with similar characteristics are grouped together.
David Jeffery, associate professor of wine science at the University of Adelaide (Australia) and co-author of the book Understanding the Chemistry of Wine, said the new technology will also have many applications in agriculture and food production.
According to him, a range of factors, from grapes, soil to microclimate and winemaking process, all influence the concentration of compounds in the finished product.
Professor Pouget shares a similar view. He says the new method could be used to monitor quality throughout the winemaking process and ensure wines are well blended.
To this day, this is done by a handful of winemakers. The pay for these people is often very high and directly proportional to their level of skill.
Meanwhile, Neuroscience page quoted professor Serge Belongie, an AI expert at the University of Copenhagen (Denmark), says training computers on human sensory experiences will create better algorithms and benefit users.
Source link
Comment (0)