Feel like listening to sad songs? New 'DJ-like' algorithm creates playlist, plays music that matches your moods
The programme identifies not only the songs, but also their order.
- The Obamas spent summer listening to these songs; how many made it to your playlist?
- Hackers break into smart home, play inappropriate music & communicate with residents via camera
- The 'Gully Boy' effect: Bangalore Poetry Festival to feature a session on rap music
- Magic of melody: Music deemed more effective that sedatives for lowering anxiety
"Whether you're getting into the car after a long day of meetings, or you're getting out of bed on a weekend morning, it should tailor its recommendations to your changing moods," said Maytal Saar-Tsechansky, a professor at the University of Texas (UT) at Austin in the US.
The programme runs a series of feedback loops. It tries out a song, the listener rates it, and the programme heeds that rating in choosing the next song.
"Then you alter the model accordingly," said Elad Liebman, a PhD student in computer science at UT Austin, who conceptualised the programme.
DJ-MC adapts to the listener's mood, considering not only which songs they will enjoy, but also in what order, the researchers said.
Songs are organised intelligently, leading to an expressive,"DJ-like" sequence, instead of a random, arbitrary-sounding one, they said.
According to the researchers, DJ MC plans its moves 10 songs ahead, like a chess player.
While one song is playing, the programme generates tens of thousands of possible sequences, and it predicts which one will please the listener the most, they said.
The programme serves up the next song on that playlist -- and while that song is playing, it creates and tests new sequences.
The researchers noted that in machine learning, the mechanism is known as a Monte Carlo search which inspired the name of the programme: DJ-MC.
The programme could be adapted to other kinds of media, from news stories to videos, they said.
"Learning algorithms don't have taste, they just have data. You can replace the dataset with anything, as long as people are consuming it in a similar fashion," Liebman said.