Recommending music in streaming services


I really enjoyed this project, reviewing how recommendation systems work throughout music streaming services (Spotify, Apple Music, etc). My early notes:

  • Collaborative Filtering in a nutshell: If user A & B have similar preferences, then songs liked by A but not yet considered by B will be recommended to B.
  • Performing Collaborative Filtering is based on Matrix Factorization.
  • Content based filtering is used often in such services.
  • The basic concept in content based filter: compare sets of audio signals which represent items in a meaningful way.
  • Spotify does this by comparing audio signals. E.g. Spotify radio.
  • Deep learning method also used (machine learning).
  • Deep learning method “mimics the architecture of mammalian brains”. It can learn features at multiple levels from low-level data without resorting to manually crafted features.
  • Deep learning “strives to provide a model for human cognition”.


  • Semantic issues in music recommendation systems…
  • Various properties cannot be obtained from audio signals.
  • Characteristics which influence user preference are not equal to the corresponding audio signal.
  • Item & user vector shortcoming or the Harry Potter effect.
  • Popular items in a recommender can lead to feedback loops. Rich get richer. If Queen gets purchased by more users. Likely to be listened by more users. Likely to be recommended to more users. Harry Potter effect lessens the diversification of items.


Exploring musics affect on combat sports

A few thoughts of mine before I wrote an essay on sport performance & music on a Music Psychology course. They’re just thoughts with little empiricism at this stage. I enjoyed these thoughts.


  • Music & dance have evolved together.
  • If it’s true that fighting is an extension of human movement, then correlating music as a stimulant for combat movement could be a powerful force that is relatively unexplored.
  • In a gym setting, music is often chosen as a supplement to the physical task at hand.
  • My feeling is that in a gym (recreational, boxing gym, MMA gym, etc) music is chosen as a mood regulator.
  • But I don’t feel this undermines music in a sporting context.
  • Rather it makes music utilitarian in this context. Music’s purpose is clearly defined if it is restricted solely to mood regulation in this context.
  • Mood regulation > emotional regulation > motivational > preparation


  • Authors Paul & Steinlage write: music became “part fight-manual, training the participant to use sound and rhythm as patterned corporeal engagement.”
  • This is evidence of embodiment.
  • They also write… “music selection is used to psychologically prep a fighter.”

Discussion 1, Genre & Gender

  • Music selection is often gendered. I.e. Males & Females tend to want different content within their music. This will naturally impact performance.

Discussion 2, The Role of Music & fighting

  • Rhythmical qualities of music lead to an increase in energy and physical output in the few studies analyzed…
  • What are the neurological implications?
  • Dopamine release > peak pleasure > increase in physical exertion
  • This is a condensed breakdown for peak-pleasure, peak-experience.
  • Should the music aim to produce peak experiences for the participants to increase athletic performance?