AT&T Labs - Research

August, 2009. Please see here for details about our exciting run to the Netflix Grand Prize - combining with Pragmatic Theory and Big Chaos to win the $1M Grand Prize!

Previous updates about our team:

December 10, 2008 We are proud to be the recipient's of Netflix' 2008 Progress Prize!. This year, we combined with our colleagues at Commendo Research in Austria to get to a 9.44% improvement over Netflix' Cinematch algorithm.

November, 2008 Our BellKor Team was featured in a recent issue of the New York Times Magazine! Read the article online or check out the accompanying video.

March 21, 2008: An update on our progress in the Netflix Prize.

Team BellKor/KorBell - Netflix Progress Prize Winners

Team BellKor is made up of Bob Bell and Chris Volinsky, from the Statistics Research group in AT&T Labs, and Yehuda Koren, who recently left AT&T Labs for Yahoo! Research in Israel. Our team has won the first two $50,000 Progress Prizes awarded by Netflix as part of their $1 Million competition to improve their recommendation algorithm. Check out their current performance on the Netflix leaderboard, and watch this AT&T Tech Channel interview with researchers Bob Bell and Chris Volinsky.

With the best score at the one year anniversary of the competition, they won the coveted Progress Prize. Netflix sent a cool commemorative plaque, although at a square meter in size, and weighing almost 200 pounds it is more like a monument!

Papers describing our techniques in analyzing the data and our experiences in the competition:

The BellKor 2008 Solution to the Netflix Prize. This is the document which lays out our overall strategy - as was required in the rules of the competition in order to claim the Progress Prize.
Factorization Meets the Neighborhood: a Multifaceted Collaborative Filtering Model. KDD 2008..
Recent Progress in Collaborative Filtering. RecSys 2008
Factor in the Neighbors: Scalable and Accurate Collaborative Filtering. submitted
Chasing $1,000,000: How We Won The Netflix Progress Prize. ASA Statistical and Computing Graphics Newsletter. Volume 18, Number 2.
Lessons from the Netflix Prize Challenge. SIGKDD Explorations, Volume 9, Issue 2.
The BellKor Solution to the Netflix Prize. This is the document which lays out our overall strategy - as was required in the rules of the competition in order to claim the Progress Prize.
Scalable Collaborative Filtering with Jointly Derived Neighborhood Interpolation Weights. ICDM 2007.
Improved Neighborhood Based Collaborative Filtering. KDD 2007 Netflix Competition Workshop.
Modelling relationships at Multiple Scales to Improve Accuracy of Large Recommender Systems. KDD 2007 .

See some more detail about our Netflix entry here

Find press coverage here

Thank You

We would like to thank Netflix for putting on a heck of a competition. The nature of the competition and especially the collaborative spirit of the participants has made this an exciting and rewarding scientific endeavor. We also would like to thank all of our competitors on the leaderboard for pushing us harder and forcing us to come up with the best possible solution. We look forward to chasing the million bucks with you all!

Please email us at bellkor@research.att.com



Legal Notice © 2011 AT&T Intellectual Property. All Rights Reserved. AT&T and the AT&T logo are trademarks of AT&T Intellectual Property.