Visualizing The Movement Of Music Genres

Music genres are challenging. In this streaming era, genres are becoming more blended, and even Spotify has labeled over 3,000 (sub)genres to their available content.

I was still curious to explore the possibility of visualizing genre metadata, in order to get a better understanding of potential movements in popular music.

Because I’m from The Netherlands I decided to focus on the Dutch Spotify charts, that echo popular music in The Netherlands, and see how visualizing genre data could help us understand its music landscape with publicly available data.

Data

Before being able to visualize anything I had to look for the right data. Spotify’s charting website provides a daily top 200 ranking for songs and their accumulated streams for that particular day. By scraping this website I’ve gathered three years’ worth of daily chart inclusion data.

For each artist that is included in that dataset, I consulted Spotify’s API to get their (by Spotify) assigned genres.

I then had to create a variable that assigned if an artist was either a local or international artist. With local music, I mean artists that have ties to The Netherlands, regardless of them singing in Dutch or another language.