Features

GPS: Question 2: Tell our readers about how you got involved in the Netflix Prize?

Simon Funk: Right around the time I was climbing on a plane for my trek around New Zealand (i.e., in some sense the worst possible time!), I got this enthusiastic email from my pal Vincent:

It looks like someone designed the contest of your dreams.

All you need to develop is the guts of the thing--no UI, no communications stuff, no infrastructure, no ancillary programs, no integration, no hardware or hardware interfaces. No need to recruit, pay, or work with a team. No tedious grant writing. No fund raising. No PR.

It looks like a pure pattern recognition problem that may be squarely in the path of where you want to go. Based on what I've seen, I think you could do this using some of the work you've already done.

There was more, but you get the idea. Honestly, I looked over the project and decided the odds of us actually winning were tiny (I still believe that), but since collaborative filtering is a big component of some of the projects I have in development, this seemed like a great data set to hone those ideas on.

What's this?