Google Reportedly Recycles Amazon Algorithm for YouTube

    February 3, 2011
    Chris Crum

Google blasted Bing in an official Google blog post for tapping into its search results. We covered that here

In that post, Google’s Amit Singhal said, "At Google we strongly believe in innovation and are proud of our search quality. We’ve invested thousands of person-years into developing our search algorithms because we want our users to get the right answer every time they search, and that’s not easy. We look forward to competing with genuinely new search algorithms out there—algorithms built on core innovation, and not on recycled search results from a competitor."

Apparently recycling is ok when it comes to recommendations, however. An interesting post from Amazon recommendation engine creator Greg Linden has come out claiming that Google-owned YouTube is using an old Amazon recommendation algorithm for video recommendation. Linden reports:

Greg Linden talks YouTube use of Amazon AlgorithmIn a paper at the recent RecSys 2010 conference, "The YouTube Video Recommendation System" (ACM), eleven Googlers describe the system behind YouTube’s recommendations and personalization in detail.

The most interesting disclosure in the paper is that YouTube has switched from their old recommendation algorithm based on random walks to a new one based on item-to-item collaborative filtering. Item-to-item collaborative filtering is the algorithm Amazon developed back in 1998. Over a decade later, it appears YouTube found a variation of Amazon’s algorithm to be the best for their video recommendations.

Google is toying around a lot with recommendations in its regular Google search results. They just started incorporating Hotpot results into the main search results. It’s unclear, however, just how many people are actually using Hotpot. As I referenced earlier, Google also recently acquired fflick, which offers opinions from users’ Twitter friends on movies (and was reportedly set to expand to more verticals). The fflick team went to YouTube, and according to a recent report from Bloomberg Business week, the upcoming "Google+1" social layer "is designed to cull data about relationships among users from current services such as Gmail and YouTube."

"Google will then let users share material through those connections, while using the information to make other products more social," the report said. "Search results may be skewed toward pages that your friends found useful."

Whether or not the Amazon-style recommendations will enter Google’s search equation is a mystery, but we thought the timing of Linden’s post was interesting, considering Google’s words about algorithmic innovation. To be clear, Linden tells WebProNews that the timing of the report was coincidental.