Facebook Updates Its News Feed Algorithm To Target Clickbait
Like art, clickbait can be hard to define, but people generally know it when they see it. Facebook also realizes that its users tend to dislike that type of content, so it's tweaked its algorithm once again to cut down on the number of clickbait headlines people see in their News Feeds.
It's not the first time Facebook has made such changes. Two years ago, for instance, it began factoring in how quickly users returned to Facebook after clicking news links; the faster people came back, the logic went, the more likely the headline was to be clickbait for content that really wasn't all that interesting.
This time around, Facebook analyzed tens of thousands of headlines to identify two key clickbait attributes: withholding information and creating misleading expectations for readers. Based on that analysis, the social media giant is now updating its News Feed system to look for common phrases associated with such attributes and filter out headlines using that language.
Aiming at Time-Wasting Headlines
"We've heard from people that they specifically want to see fewer stories with clickbait headlines or link titles," research scientist Alex Peysakhovich and user experience researcher Kristin Hendrix wrote yesterday in a Facebook news post. "These are headlines that intentionally leave out crucial information, or mislead people, forcing people to click to find out the answer."
Peysakhovich and Hendrix noted that, while the clickbait-targeting changes made in 2014 helped reduce the number of headlines similar to "You'll Never Believe . . ." that Facebook users were seeing, those changes didn't eliminate the problem.
"[W]e're still seeing Pages rely on clickbait headlines, and people are still telling us they would prefer to see clearly written headlines that help them decide how they want to spend their time and not waste time on what they click," they said.
So Facebook put a team to work to identify additional markers of clickbait to further reduce the appearance of such content. The new system being rolled out will work in much the same way as email spam filters, the Facebook researchers said.
Goal: 'Meaningful and Informative' Content
Facebook has already made several updates to its News Feed algorithm this year to address a number of different issues. In June, for instance, the company made changes that emphasized content from friends and family over other posts, noting that the goal was to ensure users saw stories that were meaningful, entertaining and informative.
That update also came in the wake of complaints of bias from some political conservatives after a Gizmodo article reported that Facebook's human news curators regularly avoided sites like The Blaze and Breitbart.
"We are not in the business of picking which issues the world should read about," Adam Mosseri, vice president of product management for News Feed, noted at the time.
The latest changes to News Feed won't result in any significant changes to the way content is distributed on most Facebook Pages, Peysakhovich and Hendrix noted in their post yesterday.
"However, Web sites and Pages who rely on clickbait-style headlines should expect their distribution to decrease," they said. Publishers can avoid such results by following Facebook's recommended best practices for News Feed content and avoiding headlines that exaggerate or mislead.
"We will learn from these changes and will continue to work on reducing clickbait so News Feed is a place for authentic communication," the researchers added.
Image credit: Facebook ; iStock/Artist's concept.
Posted: 2016-08-07 @ 6:39am PT
Posted: 2016-08-06 @ 5:44am PT
MR is an excellent example of clickbait. In addition when a headline says see the top ten . . . then has you have to click ten different times, it's also an extreme example of wasting peoples time. Most, like myself WILL NOT click more that once.
Posted: 2016-08-06 @ 4:40am PT
This is a wonderful attempt comprising a long needed moral harness on lying propagandist/journalists, and moving the technology of Artificial Intelligence ahead at the same time. Only problem is...if the methods work, basically ALL current headlines will trip the switch and the content will be blocked. Mercy, me oh my, where will we turn to fulfill our blind needs for unfiltered lies then?