[OC] Persönlicher Spotify-Verlauf der letzten 9 Jahre nach Genregruppe – Die Hörhäufigkeit ist proportional zur Farbintensität

Von Poa_Alpina

3 Comments

  1. Poa_Alpina on

    source: personal streaming history/ [https://github.com/Conchobhar/spotify-unwrapped/](https://github.com/Conchobhar/spotify-unwrapped/)

    tools: python/ seaborn

    Quick summary of technical steps:

    * Request and download Spotify streaming data
    * Get artist genre metadata from Spotify
    * Embed genres with word2vec (treat each artists genres like a sentence, but with a context window the size of the sentence)
    * Cluster genre embeddings (kmeans) into ~30 groups
    * Automatically name each cluster based on the top 2 genres in each cluster by listen count

  2. conspiracie on

    As a former Irish dancer… how the hell are you listening to that much jig and reel music without losing your mind

  3. Great work! I just requested my complete Spotify history, haha. Didn’t know you could do that. They say it takes a month to get it, though. Which is annoying because I want to use your tool now!

Leave A Reply