Facebook rolls out tweaks to combat fake news in News Feed

Facebook Icon

Facebook’s stepping up its efforts to combat fake news with two major tweaks to its News Feed algorithm. The social media giant has been the recipient of harsh criticism for the past few months for the role it was accused of playing in helping spread bogus information.

Facebook’s changes will help better rank posts in a person’s feed and show content that is most relevant to them. Its basing these modifications on the feedback it got from users who preferred authentic stories that were considered genuine and not misleading, sensational or spam-leaning.

As a result, Facebook says it’s added new universal signals to help its system decide whether or not a post is the real deal. The company went about doing this by categorizing Pages to investigate if they were putting up spam or trying to manipulate the existing system by asking for likes, shares and comments.

Also Read: Facebook announces Journalism Project to curb fake news

The three are important barometers for Facebook’s present algorithm to suss out how high to rank a post in a member’s News Feed. After identifying the guilty party, the brand utilized posts from these Pages to train a model which is supposed to single out content from other Pages that are probably authentic.

For instance, if a Page’s posts are continuously being hidden by readers, then that is taken as a sign of falseness. Moreover, bona fide content may show up higher in a user’s News Feed. Facebook is also planning to take into account how signals vary in real time when trying to bear out the relevancy of a post.

Paying attention to real-time signals means that posts about things like an individual’s favorite basketball team winning a game appears higher up in News Feed. Facebook’s reasoning behind this is that many users will be talking about it on the site for a particular length of time, making it temporarily more relevant.

Facebook doesn’t expect these News Feed tweaks to significantly alter a Page’s distribution. It expects some to get a small rise in referral traffic and others to see minor falls.