Facebook to Debut Three-Pronged Plan to Combat Fake News

Facebook issued a request for proposals from academics to study fake news on the social platform’s News Feed, with the aim of getting more information regarding the volume of false news and its impact. Those academics whose proposals are accepted will be funded and have access to data on the site. Facebook also plans to introduce a public education campaign about what fake news is and how users can stop spreading it; the campaign will be placed on the top of Facebook’s homepage. The company will also debut a 12-minute video about fake news.

Wired had exclusive access to News Feed staff who revealed the three-pronged approach. The company seems to be most enthused about the video, “Facing Facts,” which “stars the product and engineering managers who are combating false news, and was directed by Morgan Neville, who won an Academy Award for ‘20 Feet from Stardom’.”

Facebook_Apple_TV_Stream

Similar to Neville’s documentary, which focused on back-up singers, “Facing Facts” is also “told through close-up interviews and B-roll of his protagonists staring pensively at their screens.”

In addition to highlighting the work Facebook is doing to end fake news, the video is an apology, and intended to convey that the company is now “committed to cleaning it up.”

“It was a really difficult and painful thing,” said Adam Mosseri, who formerly ran News Feed. “But I think the scrutiny was fundamentally a helpful thing.”

Among those interviewed in the film is News Feed’s new head John Hegeman, who “helped build the Vickrey-Clark-Groves auction system for Facebook advertising, which has turned it into one of the most profitable businesses of all time.”

To combat the problem, Facebook is also “integrating systems — used by Instagram in its efforts to battle meanness — based on human-curated datasets and a machine-learning product called DeepText.”

First, humans identify and classify clickbait, and then machine-learning algorithms go to work, learning word patterns of what is considered clickbait and analyzing “the social connections of the accounts that post it.” The algorithm ultimately is as accurate as humans, and, importantly, much faster. Applying this to fake news is more complex; product manager Tessa Lyons notes that, “truth is harder to define than clickbait.”

To solve that problem, “Facebook has created a database of all the stories flagged by the fact-checking organizations that it has partnered with since late 2016, [and] then combines this data with other signals, including reader comments, to try to train the model.” The system also looks for duplication because, said Lyons, “the only thing cheaper than creating fake news is copying fake news.”

Facebook is “working hard on the problem,” says Wired, because fake news is “only going to get more complicated, as it moves from text to images to video to virtual reality to, one day, maybe, computer-brain interfaces.”