Similar-song finder from audio features
Inspired by publicly reported work at Spotify
Find songs that sound alike from their audio features, and group them into moods a listener would recognise.
The brief
Spotify has written about personalising its home screen with machine learning. Start with one building block: from public track data (tempo, energy, danceability, acousticness and genre), find songs that sound like a chosen song and cluster tracks into moods. Good looks like: sensible neighbours for test songs that a listener panel agrees with, named clusters, and a small app to explore them.
Read the case first
What Spotify has said in public about this kind of work, on its own pages:
- For Your Ears Only: Personalizing Spotify Home with Machine Learning (Spotify Engineering) ↗
- Spotify Research ↗
Make it your own
Tune it to listeners you know: Indian film music, devotional songs or study playlists. Decide which features matter for them and show where audio features alone get it wrong.
AITeky has no link with Spotify, which has not reviewed or endorsed this brief. The brief only draws on what the company has published, uses open data rather than its data, and is a learning exercise: don't present your project as Spotify's work or system.
What you will build
- Clean the track data and remove duplicates
- Scale features and find nearest neighbours
- Cluster tracks and name each mood
- Check suggestions with five listeners
- Build a "more like this" explorer
Free, open data to build it with
Sign up free to get the data and links
It takes a minute with your email. AiTeky Pass (₹999/month or ₹9,999/year) is free for your first 30 days.