The songs that are hits around the world today have a great deal in common, especially in the pop genre. These songs are usually built around four common notes. But of course, not every song that uses these four notes is equally successful.
Likewise, whether we like a song or not may depend on the mood and environmental conditions we are in at that moment. However, it is scientifically possible to predict whether a song will become a hit or not.
Our brain and artificial intelligence can enable us to predict hit songs:
Three scientists have revealed in their new research that we can use artificial intelligence to predict whether a song will become a hit. Of course, the song is not played to artificial intelligence here.
The reactions seen in the bodies of the volunteers listening to the song are analyzed by machine learning algorithms. Algorithms measure and evaluate one’s emotion and attention to music.
The longer a person is neurologically “in” into a song and the less they want to drop it, the more likely it is to become a hit. In other words, the person does not simply say “I like” or “I don’t like” with his mouth. His body says it instead.
In studies conducted with 33 volunteer participants, it was seen that this system was quite successful. A total of 24 new songs, 13 of which were presented as “hits” by music platforms, were played to the participants. Participants were not told which songs were hits.
During the listening sessions, the neurological responses of 33 people to the songs were transferred to the algorithms. The algorithm was able to predict which song was a hit or not, with 82% accuracy, from the neurological reactions of people by listening for just one minute.
The next focus of scientists is to predict whether entertainment content such as movies and TV series will be popular, not just music.
54488
