2. songprof.fr:
Proof-of-concept recommendation engine
hypem.com user data
HypeMachine does not make recommendations.
Vs. competition / lost engagement
Users ♥
songs
they like
Blog posts
aggregated
into
play-list
Hype Machine:“The best place to find new music on the web. ”
3. The data:
125,566 Songs
9,000 Users
900k User + Song
Interactions to model
An example of 687 songs with
27 Users in common
(Limited to interactions within last year)
Bipartite Graph relationship
4. Commonalities in behavior -> song recommendations
Collaborative filtering
Method:
1) Generate features based on nearest neighbor song commonalities, between users
2) Use ML to classify quality of recomendations, based on features from 1)
Why use two stage approach?
Network properties alone not sufficient for predicting if song liked.
Optimal network features need to be discovered.
6. Song
User
User
♥
User
User
♥
♥
Song
Song
Song♥
Possible
Recommendation
User
User
Training/test sets: 67k/18k songs
♥ = Number of users that both liked a song
Total number possible likes
Gradient Boosting Machine (GBM) Classifier. 6 fold CV.
Did user like song, yes/no?
78 % Raw accuracy (liked yes/no)
64% Predictive accuracy (AUC)
Out of sample validation of model:
2) Use ML to classify quality of recomendations, based on network features
For subset of data, generate predictions,
check against historical ♥ data:
7. Demo:
www.songprof.fr
Songs ordered by likelihood of being liked,
Given source song being liked
Based on network similarity metrics
Fed into model
Trained on actual‘likes’behavior