Get the latest tech news
BBC – Organizing music for organic discovery
Details on how we organised thousands of tracks to build a music discovery service that lets you find your next favourite song from BBC Introducing acts.
It provides machine learning models to extrapolate musical information from audio files (previously used in another project to analyse the Introducing database) such as danceability but also how aggressive, happy, sad or relaxed a track sounds. To strike a balance and still distribute the tracks on a 2D plane, we used a dimensionality reduction technique popular in data visualisation: a Principal Component Analysis (PCA). — Wreckship, BBC Introducing act The PCA gives us x/y coordinates for each track and we can use the distance between these points to determine their musical similarity — much like in Paul’s paper, but in two dimensions and with different indicators.
Or read this on Hacker News
