Music streaming services are upping the ante on their track recommendations


Is there any hope of escaping the music recommendation algorithms of Apple Music, Deezer and the likes? Yes, but that requires some effort, according to Knees. — AFP Relaxnews

As music streaming platforms draw in more and more subscribers with their millions of tracks, Spotify and the likes are striving to create advanced recommendation systems to match individual user tastes as closely as possible. But at what price?

Back in the day, if you wanted to listen to non-stop music, you needed a hefty stock of records, cassettes or CDs to hand. Now, you don’t need any of that, as music streaming platforms like Deezer and Apple Music offer “virtually infinite” musical catalogues.

In fact, assuming that the average song lasts around three minutes, it would take some 285 years to reach the end of the 50 million tracks available on Spotify. But that isn’t likely to happen. Built-in recommendation services make sure that you always have something to listen to and, above all, that these suggestions match your musical tastes.

While these algorithms draw on all kinds of information to do their job, Spotify was recently granted a patent for technology allowing its algorithms to take into account users’ ways of speaking. The technology would extract metadata from our conversations relating to our emotional state, age, gender and even our accent, in what the Swedish streaming giant describes as “an entirely different approach to collecting taste attributes of a user”.

It could also gather information about our musical preferences from “sounds from vehicles on a street, other people talking, birds chirping, printers printing, and so on”.

So how exactly could this technology be transposed into Spotify applications? While the answer to that question is not yet entirely clear, this technology is testimony to the levels of creativity and inventiveness being deployed by streaming sites to make sure you're always listening to just the right track. And it turns out that streamers have a relatively short amount of time to tap into users' thirst for discovering new music.

A study from Deezer revealed that people are inclined to discover new artists and songs up to the age of 30 years and six months. After that, they enter a kind of “musical paralysis”, preventing them from venturing beyond their usual playlist. And streaming giants are increasingly taking into account this lack of curiosity in their own business model.

And that's exactly where playlists come in, these musical selections that can be tailored to our tastes and to all kinds of situations. Whether it’s a soundtrack for cooking pasta or delving into the fascinating world of sea shanties, there are playlists for both of those things.

Playing the algorithm game

Certain playlists, like RapCaviar, Viva Latino and Today's Hits, are followed by millions of music lovers, and can bring a considerable revenue boost to artists whose tracks are featured in the selections.

According to a study from the EU’s Joint Research Centre, getting to the top of Spotify’s New Music Friday playlist was potentially worth up to US$117,000 (RM474,844) to artistes in 2018 – a figure which has probably increased since. As a result, in November, the Swedish streaming platform announced a new function that would give artists a boost in its algorithms... if they accepted to hand over a portion of their royalties.

The initiative proved particularly controversial in the industry, and led the British government to launch an inquiry into the business model of music streaming.

So is there any hope of escaping the music recommendation algorithms of Apple Music, Deezer and the likes? Yes, but that requires some effort, according to Peter Knees, assistant professor at TU Wien – Vienna University of Technology.

“When you keep listening to the recommendations that are being made, you end up in that feedback loop, because you provide further evidence [to the algorithms] that this is the music you want to listen to, because you're listening to it," he explains to Wired.

Certain platforms like Bandcamp claim to have a more “human” approach to music recommendations, offering users the opportunity to discover new music that flies “under the radar” of algorithms. All of which just goes to show that you’ll never likely run out of tracks to feed your hunger for music. – AFP Relaxnews

Article type: metered
User Type: anonymous web
User Status:
Campaign ID: 1
Cxense type: free
User access status: 3
   

Next In Tech News

Tesla says sorry to Chinese buyers, back-pedalling on its ‘no compromise’ attitude towards ‘unreasonable’ customer grievances as pressure mounts on social media and state press
Music and tech team up to cut the milliseconds
Amazon to let Whole Foods shoppers pay with a swipe of their palm
Instagram launches feature to tackle hate speech, abuse
Tesla protest at Shanghai Auto Show 2021 ends with woman dragged off by security after climbing onto car and shouting
Selangor cops warn of online scammers offering jobs as e-commerce purchasing agents
Two die in driverless Tesla incident – where are the regulators?
Top online brokerage lures young Japanese with zero-fee trading
ASML Q1 profit beats estimates, order book surges amid semiconductor shortage
Worldline reports decline in Q1 sales on health curbs

Stories You'll Enjoy


Vouchers