• by Tobi Elkin  @tobielkin, May 11, 2017

In a bid to make it more challenging for publishers to monetize low-quality Web sites, Facebook will use artificial intelligence to locate sites that have too many ads or ads that are disruptive or malicious-looking.

This move is part of the company’s plan to show consumers “fewer misleading posts and more informative posts,” resulting in more traffic for publishers of quality content.

Facebook said it determines what a low-quality site is by looking at the ratio of ads to actual content. For example, a page with too many ads and not enough original text or images is considered spam.

If a site's ads are "disruptive, shocking or malicious," Facebook will make sure fewer people can see and click on it. While the site links aren’t banned -- if users post a link, it will still appear on their wall -- the poster's news feed ranking will decrease so the feed will be seen by fewer people.

The changes will take place over the next few months.

Facebook said since last year it has been banning "advertisers with low-quality Web-page experiences from advertising on our platform."

Facebook offered advice for companies concerned their traffic will drop: They should keep a watch on the number of ads on a page. Sites may be flagged if ads have overtly sexual, violent, or scary images, as well as ads promoting illegal products or contain malware, and pages with too many pop-up and interstitial ads.