Home 🏠
Automated Music Recommendation · v2.3
MIT Media Lab · 1993–1996
Community: 847 active listeners
Songs rated today: 12,041
RINGO was one of the first collaborative filtering systems (MIT Media Lab, 1994). Unlike content-based systems that match song features, RINGO finds other users whose tastes match yours — then borrows their ratings for songs you haven't heard. Rate songs 1 (hate) → 7 (love) and watch the algorithm recalculate in real-time. The more you share with a neighbour, the more their opinions shape your recommendations.
RATE THESE TRACKS
YOUR RECOMMENDATIONS
Rate at least 3 songs to generate recommendations…
NEAREST NEIGHBOURS
Rate songs to discover similar listeners…
COMMUNITY HEAT MAP
SCORE BREAKDOWN
Each recommendation score =
▸ Neighbour votes     60%
▸ Community average   25%
▸ Genre affinity      15%
Neighbour weight is Pearson correlation between your ratings and theirs — users who agree with you on rated songs get more say in unrated ones.
RATINGS: 0
NEIGHBOURS FOUND: 0
RECOMMENDATIONS: 0
COVERAGE: 0%