Facebook Targets Low-Quality Links With Spam-Fighting AI
Fake news isn't the only objectionable content the world's largest social network is trying to stamp out.
Facebook is rolling out new policies, backed by artificial intelligence algorithms, that will sniff out and remove from its users' news feeds links to sites with annoying and misleading ads. Facebook says it has always prevented advertisers with "low-quality web page experiences" from buying ads, but it will expand the ban to spammers who post organically so their links show in news feeds.
"Starting today, we're rolling out an update so people see fewer posts and ads in News Feed that link to these low-quality web page experiences," the company wrote in a Wednesday blog post. "Similar to the work we're already doing to stop misinformation, this update will help reduce the economic incentives of financially motivated spammers."
There are two components to the new spam-fighting tool. The first is a database of hundreds of thousands of web pages that contain "disruptive, shocking or malicious" ads and little substantive content. These are the same type of websites Google punishes in its search results rankings.
The second component is an AI algorithm that scans new posts on Facebook to determine if they match the types of spam in Facebook's database. If the algorithm detects a high match for a specific post, it will push the content lower down in people's News Feeds. If it detects a high match for an ad, that advertiser may be banned.
"We hear from our community that they're disappointed when they click on a link that leads to a web page containing little substantive content and that is covered in disruptive, shocking or malicious ads," Facebook wrote. "People expect their experience after clicking on a post to be straightforward."
Don't expect a huge decrease in news feed spam right away, though. Facebook says it will deploy its new algorithm gradually over the next few months.