Facebook Surveys Users On News Story Relevance


Facebook is polling users on how informative they find stories linked to on the site. It’s another attempt to cut back on clickbait.

The move follows several recent changes to the algorithm for the News Feed, the default view on Facebook by which status updates, links and other content appears in a perceived order of relevance to the user rather than in chronological order.

In the past few weeks Facebook revealed it is putting extra emphasis on content posted directly by a user’s friends, rather than third party content the friends have shared. It’s also made changes, albeit unexplained ones, designed to downplay links where headlines don’t accurately reflect the content of the linked article.

The company has now revealed details of a Feed Quality Program by which “tens of thousands of people per day” are asked to rank stories in their news feed with a score from 1 to 5 for relevance. Facebook’s algorithm will then try to figure out common characteristics among those stories with the higher ratings and use these criteria to make assumptions about how generally relevant new stories will be.

Facebook says it will then combine this general relevance ranking with factors specific to the individual user. For example, if a user regularly comments on or shares stories from a particular publisher, they should be more likely to see future stories from that source. Similarly, if a user regularly interacts with stories shared by their work colleague but usually appears to skim past and ignore stories shared by their aunt, this will also affect future rankings.

Geeks are Sexy needs YOUR help. Learn more about how YOU can support us here.