Complaints rang far and wide last week after an automated system for spotting inappropriate Facebook comments began blocking legitimate posts by the social network’s users, including prominent members like Robert Scoble.
Facebook users of all stripes have received warnings about posting “irrelevant” or “inappropriate” comments and some have had comments blocked outright in recent weeks. However, the company has been mum on the cause of the incidents.
Reports of blocked comments on Facebook began popping up in recent weeks. But the practice reached the attention of the media after Mr. Scoble had a comment run afoul of the site’s filters on Saturday. Scoble, who runs the Sobleizer blog, is a leading expert on social media who works for the hosting firm Rackspace and enjoys a massive and influential online following. He blew the whistle on the filtering in a post to his Google+ account, after his attempt to comment on a discussion about the news site PandoDaily was blocked and labeled by Facebook as “irrelevant or inappropriate” and not contributing to the post “in a positive way.”
“Has anyone ever seen anything like this before? I haven’t, and I’ve posted tons of comments to Facebook,” Scoble asked in a post on his Google+ account.
Scoble wasn’t the only user to encounter problems. Jarin Udom, a Facebook user and founder of the mobile and Web application development shop Robot Mode in San Diego, California, had an innocuous comment posted on a friend’s wall about the weekend “Super Moon” flagged as inappropriate. Udom said he still isn’t sure why, as his comment didn’t include links to other Web pages or inappropriate language of any sort.
Scoble reported that he was contacted by a Facebook PR representative who said that the flagged comment was possibly a “false positive” from the company’s comment spam filters. A company spokesman told Threatpost that the company could not immediately comment on the issue.
The rash of strangely flagged comments may reveal the strain of Facebook’s recent efforts to crack down on both spam and malicious links, and its efforts to identify and stamp out online bullying using its platform.