The second Netflix prize is canceled due to privacy problems. I continue to believe my original assessment of this paper, that the privacy break was somewhat overstated. I still haven’t seen any serious privacy failures on the scale of the AOL search log release.
I expect privacy concerns to continue to be a big issue when dealing with data releases by companies or governments. The theory of maintaining privacy while using data is improving, but it is not yet in a state where the limits of what’s possible are clear let alone how to achieve these limits in a manner friendly to a prediction competition.
One response to these kind of problems is to pull the data and require people to sign a form saying they will not try to re-identify individuals. When Homer et. al showed an attack on the genome-wide association study (GWAS), the NIH did this. Of course, this sort of defeats the purpose of Netflix’s experiment.
Interesting article that discusses changes to Netflix’ business which caused them not to implement the winning entry.
When the contest began in 2006, most customers were using the slow-turnaround DVD-by-mail system, where a bad choice meant waiting a few days. Any improvement to recommendation systems that would cause fewer “lost days” would be a needed improvement to such a system.
Now, however, the bulk of Netflix customers use the streaming interface, which allows users to change their minds in real time, so the recommender is no longer such a pressing need.
One technical tidbit in the article is a discussion of which algorithms Netflix chose to use.
See:
http://www.techdirt.com/blog/innovation/articles/20120409/03412518422/why-netflix-never-implemented-algorithm-that-won-netflix-1-million-challenge.shtml