So many ears ahead
A recent research paper published by the Tel Aviv University’s School Of Electrical Engineering makes a mockery of the manner in which record companies work or, more precisely, the ‘hit-or-miss’ successes — with the stress being on the misses, writes Parag Kamani.music Updated: Jan 13, 2009 13:13 IST
A recent research paper published by the Tel Aviv University’s School Of Electrical Engineering makes a mockery of the manner in which record companies work or, more precisely, the ‘hit-or-miss’ successes — with the stress being on the misses.
For every singer makes it to the global sales chart, there are thousands who don’t, even though they have been signed by the same label. What is this study that has made talent scouts redundant.. well, almost?
The University has devised an algorithm that spots new artistes several months before they achieve national or –– as this case is –– international success.
Using data collected over a nine-month period from Gnutella, the world’s most popular peer-to-peer file-sharing network that is incorporated in popular sites such as LimeWire, Morpheus, and Torrent, the study has led to an obvious inference.
Artistes who gain some semblance of recognition in their local neighbourhood, initially with a handful of enquiries, eventually get city-wide recognition, and then nation-wide acclaim. Through this process, the system has proved itself a reliable predictor of breakout artistes.
Just to ensure that the system really works, three artistes and their songs –— Soulja Boy’s Crank that, Sean Kingston’s Temperature, and Shop Boyz’s Party like a rockstar –– were ‘recommended’ by the University as potential hits in the U S.
The songs did make it to the Billboard chart within months of the recommendation. Artistes tipped since have ensured a 30 per cent success rate for the algorithm, with the success rate going to as high as 50 per cent in some instances. The University now hopes to predict whether songs of established artistes will succeed.
Also, on the cards, a test to predic the success of TV programmes and home video. For the moment it is enough if the software indeed predicts new music idols. It could become a profitable tool for record labels which then won’t have to sign or support artistes who have little chances of success.
It could be a reality-check,too, for budding artists to gauge whether they should pursue a career in music. So the next time you find yourself listening to a singer who could be the future Madonna or a band which could be the future U2, don’t rely on your instincts.
Rush to Tel Aviv University’s School Of Electrical Engineering for their computer algorithms to find out if the artistes in question are indeed going to be the next big thing or whether your ears need cleaning out.
First Published: Jan 13, 2009 13:03 IST