Subscribe Home/Reviews/Software/Social Networking/Facebook To Revamp Trending Topics Selection Process

Facebook will no longer rely on external web sites and news outlets to determine which topics to include in its trending news section, the company announced today.

In a letter to Sen. John Thune, chairman of the Senate Commerce Committee, the company outlined the changes that it would make to its trending topics selection process in the wake of a two-week investigation following allegations that it regularly suppresses coverage of politically conservative news topics.

Previously, the trending topics review team would rely partially on RSS feeds from online media outlets, including The New York Times, BBC News, the Drudge Report, and the Huffington Post. Going forward, it will rely only on algorithms that scan Facebook posts for an unusually high number of mentions of a particular topic. Facebook said that relying only on its users’ posts will reduce the possibility of missing a story that the media does not cover.

Besides eliminating its reliance on media outlets, Facebook also said it would make changes to its reviewer training program and add more controls and oversight. Reviewers are third-party employees hired by Accenture under a contract with Facebook.

Despite prompting changes to the review process, Facebook said its investigation did not find that the trending topics section deemphasized politically conservative viewpoints.

“The only clear trend revealed by the analysis was that moderate topics—that is, those that are popular across the political spectrum—are approved and boosted at a higher rate than liberal or conservative topics,” it said in the letter, signed by General Counsel Colin Stretch.

Sen. Thune raised concerns two weeks ago that Facebook’s use of an algorithm to assemble its trending topics section could mislead users. Following today’s letter from Facebook, he said he was satisfied with the social networking company’s actions.

“Facebook has recognized the limitations of efforts to keep information systems fully free from potential bias, which lends credibility to its findings,” he said in a statement.