LO: Computational Data Visualization

A student in CMU Design called Bryce Li created a website in 2021 that helps visualize songs by a specific artist. If you like a song by that artist, the website can show you other songs by the same artist that are similar. This information is presented in clustered data points that are visualized as circles. Each circle represents a song and the closer 2 circles are together, the more similar those songs are. When you click on a circle, it also plays a 30-second clip of the song so you can hear it for yourself. This website was coded using libraries from React and Three js. And the coding languages are HTML, CSS, and JS. To cluster different points by shared similarities, Spotify has a quantifying API that Bryce used along with a k-means clustering algorithm. I thought this was interesting because I often come across a similar situation where I like an artist’s song but I don’t know how to find similar songs. By simplifying the amount of data a user sees and taking away unnecessary information, this website, therefore, communicates clearly what the user wants to know: similar types of songs by the same artist.