Netflix offers $1 million prize for a better way to recommend movies

For those of you who don’t know, Netflix is the online DVD movie rental company that now sport 5.2 million subscribers. They are looking for ways to improve their movie recommendation system and have announced a $1 million prize for a software application that improves the accuracy of Netflix’ current system by at least 10%. To help developers, Netflix is releasing 100 million movie ratings, with all personal information removed.

Why do I post about this here, you ask? Because geography and location play a role in culture, of course, and I’m guessing that a solution that involves the locations of rental subscribers would be a pretty strong candidate. Unfortunately, that’s way beyond my skills, but some of you out there might be able to tackle such a herculean task (Clearly it’s not easy, or they wouldn’t be offering $1 million!)


Will mean more later

This won’t mean much to you until you hear the difference for yourself, but Gigavox, the business arm of IT Conversations, has released a GREAT little tool called Levelator that levels audio all the way through an audio file. To date I have only been able to do this in a patchy way through the software, but I ran it on Episode 63, which had quite a few quiet vs loud spots (eg Frank was kind of quiet this week) and the Levelator brought the audio levels between the 3 of us and the music all to the same level throughout the episode. I would go ahead and switch out the files for epi 63 if I wasn’t afraid of messing up the feed 🙂

If you do a podcast or any web audio, definitely check out Levelator …did I mention it is free…