Facebook Aims to Predict Prominence of Stories on News Feed in Algorithm Change

look at the stories that people tell us they find informative,” said Product Manager-News Feed Vibhi Kant and a pair of company scientists in a Thursday blog post. “People from our Feed Quality Program look at each story in their feed and rank it on a scale of one to five -- one being 'really not informative' and five being 'really informative.' Generally, we’ve found people find stories informative if they are related to their interests, if they engage people in broader discussions and if they contain news about the world around them.” The stories “people rate as informative and really informative help create a new prediction about how informative we think you’ll find each story,” the staff said. The website will use the new prediction signal in combination with the story's relevance data “to best predict stories that you might personally find informative,” Kant and the scientists said. “Informative stories are therefore different for each person and will likely change over time.” Most pages won't see a significant change in distribution because of the new prediction signal, but some may see either a small increase or decrease, Facebook said. The social media company has made multiple changes to its news feed algorithm in recent months following claims that it was censoring politically conservative content from appearing prominently on the feed (see 1605100032, 1606290066 and 1608040053).