Wired magazine recently published an interesting article on the Netflix Prize:
This Psychologist Might Outsmart the Math Brains Competing for the Netflix Prize
The article is a fun read. It provides some perspective on the importance of tuning algorithms and the potential for combining many algorithms for one prediction task. It also makes it clear that the prize-seeking community is very open to sharing results and techniques. Cool.
I would have been interested in reading more about why the researchers think going from 8% RMSE improvement to 10% improvement will be so hard. Is is because they've (finally) bumped up against individuals' abilities to accurately represent their own movie preferences on the 1-5 scale? I ask, because I had thought we were already there before this contest! How much room is there for algorithms to get better at predicting our individual rating idiosyncrasies and inconsistencies?
Max
Joshua Porter has written an article on the how recommendation systems are changing the Web. He lists several benefits and drawbacks stemming from these technologies. It is cool to observe that our group is actively researching to improve three of the four "drawbacks" that he lists:
- Dan Frankowski is working on Recommenders Everywhere - one attempt to make recommendation systems easier to set up
- Dan Cosley is working to providing tools that help communities build and maintain good databases
- Tony Lam is working to understand shilling in recommender systems
