IEEE named Amazon's recommendation paper the best of its 20 years

When the journal IEEE Internet Computing marked its twentieth anniversary in 2017, its editorial board reviewed everything it had published since 1997 and singled out one paper as the single article that had best withstood the test of time: “Amazon.com Recommendations: Item-to-Item Collaborative Filtering” by Greg Linden, Brent Smith, and Jeremy York, which had appeared in the journal in 2003. By then the algorithm it described had been quietly powering Amazon’s “Customers who bought this item also bought” recommendations for years, and the item-to-item approach had been adopted widely across the industry.