‘Nelson’ Allows Spotify Premium Users to Tweak its Recommendation Algorithm

If you’re a sound engineer or even an audiophile who had heard of (made-up) words like ‘Instrumentalness’ and ‘Speechiness’ before, Spotify’s latest features might just be for you. That’s because the leading music streaming service now has a new browser tool that lets Premium users fine tune the recommendation algorithm. Called Nelson, the web-based tool is still in beta, but already incorporates a number of features that power users may find useful.

According to Spotify’s Developer Advocate Engineer, Arielle Vaniderstine, (@imariari), “Nelson uses @Spotify Platform recommendations and Web Playback SDK to play the perfect music in your browser”. To get a customized recommendation playlist, all you need to do is select a few genres and the parameters that need to be changed, like ‘danceability’, ‘liveness’, ‘tempo’, etc. On its developer blog, Spotify goes into great technical details to explain how to go about things, so you can check that out by clicking through this link.

Hack of the day: Nelson. Uses @SpotifyPlatform recommendations and Web Playback SDK to play the perfect music in your browser https://t.co/d31YP7tF6V @glitch

This is a V1, feedback encouraged 🙂 pic.twitter.com/QIkhWEpJyB

— Ari V (@AriVanider) February 5, 2018

While the tool seems way too complicated for large swathes of mainstream users, some of it can be fairly simple once you play around with the settings a little. What’s more, it can definitely help you explore music outside your comfort zone. For example, if you’re a metalhead who rarely listens to American country music, Nelson will give you the opportunity to ask Spotify to recommend to you country music songs that are both fast-paced and heavy on instrumentation, since that’s what you prefer.

The ability to tweak some of the parameters of the Spotify API to get custom recommendations certainly sounds interesting, but the feature is currently in beta, and if you don’t already know what you’re doing, it will likely need a lot of trial and error to be able to find exactly what you’re looking for.