Networks between Music Genres
I just discovered a website by Glenn Mcdonald, who is the curator for Spotify’s genre based music stations. The site uses Spotify’s wealth of music data to map the different genres of music (over 1,500) onto an axis where according to the creator “down is more organic, up is more mechanical and electric; left is denser and more atmospheric, right is spikier and bouncier”.
Representing music in this way comes up with some interesting network discoveries, genres I never thought would be related (ex. Arab Pop music and Beach music). My favorite genre of music is Rockabilly, such as Elvis. Yet I also like what is called “Adult Standard” such as Frank Sinatra. According to the data plot, Rockabilly and Adult Standard are actually closely related, so my preferences are not as varied as I originally thought.
These plotted network diagrams also have interesting implications in industry. For example, in the case of internet radios such as Spotify or Pandora, music preference algorithms can take note of closely related music genres and suggest tangential artists that the user may like, even if they have never been directly exposed to the new genre.
By taking advantage of the strong ties between certain music genres, it seems that Spotify is well equipped to curate playlists for its users, based on known preferences, and deriving likely preferences, thus increasing the user’s overall satisfaction with the app. Similarly weak ties between music can indicate that the user may like the genre, but it probably shouldn’t be recommended right away. Where as no ties between the music, indicate that the algorithm is faulty if it is suggesting genres without any ties.
Source: everynoise.com
Source: https://lifehacker.com/hear-what-literally-every-genre-of-music-sounds-like-wi-1819421235