In a bid to curb the spread of misinformation, social networking giant Facebook is now expanding its fact-checking capabilities to all of its third-party analysis partners around the world to examine the authenticity of the photos and videos shared on the platform.
“We have built a Machine Learning (ML) model that uses various engagement signals, including feedback from people on Facebook, to identify potentially false content and then send to the fact-checkers for review who analyse the content metadata along with other information to assess the truth or falsity of a photo or video by combining these skills with other journalistic practices or government agencies,” Facebook said in a statement late on Thursday.
Based on testing and research, the social networking company claimed that misinformation in photos and videos falls into three categories — manipulated or fabricated, out-of-context and text or audio claim — which the platform aims to monitor and reduce.
“We use optical character recognition (OCR) to extract text from photos and compare that text to headlines from fact-checkers’ articles and now we are also working on new ways to detect if a photo or video has been manipulated,” the company added.
According to Tessa Lyons who is a Product Manager at Facebook, the company started its work on misinformation with articles because that was what people in the US said was the most prevalent form of false news they were seeing.
“What they would do is they will share articles that contain misinformation – and people would be surprised by the headlines because they were false. So, they will click on those articles and land on websites where bad actors were monetising their impressions with ads,” Lyons said.
Expanding the fact-checking capabilities of Facebook to third-party partners would add more ratings on the accuracy of the content circulating on the platform using which it would be easier to improve the credibility of Facebook’s ML model.