01
We have all found songs we love through Spotify’s Autoplay/Smart Shuffle, but how does it work? Through analyzing my music listening habits extensively, during the week of September 21st-27th, many questions were raised about how Spotify’s algorithm actually works. I was mesmerized by this concept of “energy” that Spotify creates. This value is made through several musical elements, all generated by its machine learning audio analysis. How is this algorithm understood by developers vs. consumers, and how is it presented to the world? Are there elements hidden from the general public that help identify a user’s listening habits to raise engagement? What does my algorithm look like? How can one use the algorithm as a tool? Is that even possible?
02
- Algorithmic Symphonies: How Spotify Strikes the Right Chord
-
Made to Be Found | Spotify for Artists.
-
This website is actually insane…
-
-
Inside Spotify’s Recommendation System
-
Multi-Task Learning of Graph-based Inductive Representations of Music Content
-
Is My Spotify Music Boring: An Analysis
-
Unveiling Patterns in Spotify’s Top Tracks
-
This website is also insane.. SO much data and ways of graphing it
-
-
An article about using Spotify’s algorithm to boost engagement
03
04
Personal research, I didn’t use any direct links. All information was found by manually making playlists based on listening history provided by AirBuds (app). Then, each playlist was exported using a Spotify API encoder.
05
