Facebook is now using human raters around the world, not just in the U.S., to help its algorithms decide what should appear in the News Feed.
That’s one of the key takeaways from a lengthy and interesting Slate.com article published late Sunday. The article is somewhat similar to one published a year ago on Medium, when Facebook first started talking about what it called the “feed quality panel” — a group of regular Facebook users that the company pays to provide daily feedback on content in the News Feed. At that time, the panel was about 600 people all located in the U.S. As Slate reports now, Facebook became so reliant on the group’s feedback that it expanded the panel internationally late last summer.
The article offers some interesting details about how Facebook takes what it learns from the human raters and integrates it into the News Feed:
…the algorithm is so precious to Facebook that every tweak to the code must be tested—first in an offline simulation, then among a tiny group of Facebook employees, then on a small fraction of all Facebook users—before it goes live. At each step, the company collects data on the change’s effect on metrics ranging from user engagement to time spent on the site to ad revenue to page-load time. Diagnostic tools are set up to detect an abnormally large change on any one of these crucial metrics in real time, setting off a sort of internal alarm that automatically notifies key members of the news feed team.
The Slate article is a good read and proves that Facebook has reached the same conclusion Google (and others) reached long ago: Algorithms can only do so much, and they still need human feedback to understand what humans like and why we interact with some content more than others. Google began using human search quality raters as far back as 2005, and Microsoft has had a similar program in place for its search engines (Live Search, Bing) since the same time.
And to get a better understanding of Facebook’s News Feed changes over the past couple years — including the increased reliance on user feedback — don’t miss our recent article, Facebook’s News Feed Algorithm: A Guide To Recent Changes.
No comments:
Post a Comment