Thanks to digital streaming technology, the average person now has an entire universe of music at their disposal. With such a massive library of sounds at your very fingertips, it can be hard to choose, whether you're looking for exercise music, a party playlist or that song for someone. Some people spend hours searching for the right music, only to end up wandering where the streets have no name. Streaming services already attempt to help users finds songs, but their algorithms often work in mysterious ways.
But if there were a simpler method to find songs, wouldn't that just be the sweetest thing? Now, there could be a huge breakthrough in how we find music online, and it's all thanks to rock and roll legends U2.
All because of U(2)
Sadly, Bono and the boys won't be heading to your house to help you find the perfect tunes. Instead, the iconic Irish outfit's music was used as the basis for a new study published in the Journal of Documentation.
Led by Dr. Diane Pennington from Glasgow's University of Strathclyde, the study analyzed 150 fan covers of U2's 2014 single "Song For Someone." The songs were recorded as part of a YouTube contest in which U2 asked fans to "make (it) your song." As Dr. Pennington explained in an accompanying press release, researchers recognize how important emotion is in driving people to certain songs or artists.
The only problem is, streaming services don't always recognize emotion, and their "information retrieval systems" - the complex math they use to recommend songs - look mostly at keywords. So, Pennigton and her team analyzed the U2 covers as a way to process emotions in the same way a computer might handle other queries.
So why the U2 fan renditions specifically? Because U2's music provided a "rich source" of emotional content and information, with fans conveying a number of different emotional responses. And the imagery helped, too, with fans' shirts, posters, and other signs of devotion providing valuable clues.
Pennington explained that all of this human data will go a long way to helping streaming music sites, and it could eventually lead to automatic playlists generated by emotions. Pennigton sees the system going beyond just making a sweet playlist for your pool party; doctors could use it to choose songs for music therapy.
With or without you
Pennington and Co. aren't alone in trying to bring emotion into streaming music. A team of researchers from the University of Cambridge, McGill University, and Stanford Graduate School of Business is developing three quantifiable categories to describe music. This should help services rely less on the library of search terms and adjectives.
A team from the Neotia Institute of Technology Management and Science is taking a different approach, pairing people's perceptions and emotions with certain tones. It's bit more complex than Pennington's U2-centric method, but one thing's clear: the emotions of music are an untapped resource, and we could be on the verge of breaking new ground in how music is disseminated and consumed.
Now that would make for one beautiful day.