Music recommendation is broken - automatic music recommenders make mistakes that no human would ever make. In this talk, we will explore why recommenders make such dumb mistakes and we will explore some of the new ideas coming from recommendation and music researchers to help make music recommendations better.
Questions Answered:
Why is music recommendation important?
How do current recommenders like iTunes Genius, Last.fm and Pandora work?
What are the strengths and weaknesses of collaborative filtering?
How can content-based recommendation techniques be used to extend recommendations into the long tail?
How can music recommenders be evaluated?
How can a recommender be immune to hackers and shillers?
What are some of the new ideas coming from the academic world that will help make recommenders better?
How can user interface and visualizations improve the music discovery experience?
Why are transparent recommendations (recommendation with an explanation) so important?
Why are novelty and context so important to music recommendation?
Panelists:
Paul Lamere (Sun Microsystems), Anthony Volodkin (The Hype Machine)