@serendipity
8 months ago

How does Spotify recommendation algorithm work?

Collapse
0 0
Most people love Discover weekly, So here it is -


The main ingredient in Discover Weekly, it turns out, is other people. Spotify begins by looking at the 2 billion or so playlists created by its users—each one a reflection of some music fan’s tastes and sensibilities. Those human selections and groupings of songs form the core of Discover Weekly’s recommendations.

“Playlists are the common currency on Spotify. More users knew how to use them and create them than any other feature,” said Ogle, who previously founded This Is My Jam, a startup that asked users to pick one favorite song at a time. It shut downearlier this year.
Spotify considers everything from professionally curated playlists like  RapCaviar  to your cousin Joe’s summer barbecue jams. It gives extra weight to the company’s own playlists and those with more followers. Then it attempts to fill in the blanks between your listening habits and those with similar tastes. In the simplest terms, if Spotify notices that two of your favorite songs tend to appear on playlists along with a third song you haven’t heard before, it will suggest the new song to you.
…and your own taste profile  
But the recipe for your Discover Weekly playlist is a lot more complicated than that. Spotify also creates a profile of each user’s individualized taste in music, grouped into clusters of artists and micro-genres—not just “rock” and “rap” but fine-grained distinctions like “synthpop” and “southern soul.” These are derived using technology from Echo Nest, a music analytics firm that Spotify acquired in 2014, which learns about emerging genres by having machines read music sites and analyze how various artists are described.



Collapse
Spotify uses Audio Models. Collaborative Filtering and Natural Language Processing are useful to connect listeners with music that others are already listening to and are talking about. Audio Models are useful for analyzing raw audio data and recommending new songs that have not become popular yet. This algorithm uses Convolutional Neural Networks that use clustering to identify similarities in time signature, key, mode, tempo, and loudness of audio tracks.
Collapse
Spotify doesn’t actually use a single revolutionary recommendation model. Instead, they mix together some of the best strategies used by other services to create their own uniquely powerful discovery engine.

There are three main types of recommendation models that Spotify employs: Collaborative Filtering models, Natural Language Processing models and Audio models.
Collapse
Spotify recommends songs from other peoples' playlists using two methods.
The first method is "collaborative filtering."This simply means finding other playlists that are very similar to your own and recommending you the few songs in the lists that are different.
The second method is through your "taste profile." Spotify predicts what genres and microgenres you like based on songs, artists, and playlists you repeatedly listened to in the past, and recommends new songs from similar categories. On the flip side, Spotify learns which songs, artists, and playlists you don’t like based on the songs you skip often and uses this to improve their model.
Collapse
3 comments
Not only does Spotify allow users to build their own playlists, but Spotify connects users with their friends and provides new music suggestions based on three recommendation methods that use social and music network connections and clustering. 
Collapse
Any algo that deals with personal preferences would have to do with proximity or similarity for sure.
if you have listened to songs on spotify or made a playlist in it then the algo would surely sift through all the other playlists and similar class of songs that have been heard or placed in a playlist but wouldn't have been heard by you.. hence it can consistently come up with new songs in your playlist..
Collapse
While Pandora uses content filtering, Spotify uses collaborative filtering for their Discover Weeklyrecommendation system. This latter technique uses previous users' input/behavior to make futurerecommendations. ... Content information can also be built into collaborative filtering system to improve performance.

Collapse
Through  collaborative filtering -it's implies finding other playlists that are very similar to your own and recommending you the few songs in the lists that are different.
The second method is through your "taste profile." Spotify predicts what genres and microgenres you like based on songs, artists, and playlists you repeatedly listened to in the past, and recommends new songs from similar categories. On the flip side, Spotify learns which songs, artists, and playlists you don’t like based on the songs you skip often and uses this to improve their model.
Collapse
I think its based on playlists u already have and than othet songs u play so they look for similar sounds or artists to recommend
Collapse