What wine to go with fish? Artificial intelligence will advise

If this technology continues to develop at such an alarming pace, it will soon be easier to list the issues in which artificial intelligence will not help us than those in which it will prove to be the best solution to a given problem. Unless, of course, trying to choose the type of wine for the right occasion can be called a problem.

Do you like wine? Artificial intelligence will select them for you

Although it is hard to believe, this will soon be the case – artificial intelligence will become a human assistant in selecting the right type of wine, for example, for a meal. And although mobile applications, such as Vivino or Hello Vino, which also use simple algorithms, have been helping with this for some time, it was only scientists from the Technical University of Denmark, Copenhagen and California who decided to approach this topic really seriously.

Scientists conducted special tests with 256 volunteers who had to taste a large set of various types of unsigned wines. Each of them was marked with a number and color, and the experiment participants’ task was to record information about the individual characteristics of individual wines. It didn’t differ much from typical consumer tests, but the most important thing was the next steps of the test.

In the next steps, the scientists digitized the data obtained in this way into a digital form and associated them with hundreds of thousands of labels of the most popular wines, as well as their reviews provided by users of the mentioned Vivino application, which is the most frequently used software of this type by wine enthusiasts. In this way, an advanced algorithm was created based on an almost infinite data set, which is the resources of the global leader in wine applications.

Artificial intelligence recommends the type of wine. How it’s working?

The mentioned algorithm combines data from wine labels and reviews with data from wine tastings, and then, on this basis, predicts consumers’ exact tastes regarding their preferred taste or favorite types of wine. This seemingly simple mechanism uses complex machine learning combined with researched and clearly defined preferences of study participants. Scientists from the Technical University of Denmark argue that the method they have developed is effective, and they have shared the research results online – here you can see for yourself whether this method is effective.