Facebooks Newsfeed lernt Clickbait-Headlines

Facebooks Feldzug gegen Clickbaiters geht in die nächste Runde. Vor knapp zwei Jahren führten sie einen Mechanismus ein, der die verbrachte Zeit auf den angeklickten Links maß und kürzere Verweildauern abstrafte.

Jetzt fügen sie ihrem Newsfeed ein System dazu, dass die Sprache dieser Headlines lernt. Dazu haben sie sich anscheinend einen manuell angelegten Daten-Korpus mit kategorisierten Headlines angeschafft (den ich gerne mal sehen würde).

First, we categorized tens of thousands of headlines as clickbait by considering two key points: (1) if the headline withholds information required to understand what the content of the article is; and (2) if the headline exaggerates the article to create misleading expectations for the reader. For example, the headline “You’ll Never Believe Who Tripped and Fell on the Red Carpet…” withholds information required to understand the article (What happened? Who Tripped?) The headline “Apples Are Actually Bad For You?!” misleads the reader (apples are only bad for you if you eat too many every day). A team at Facebook reviewed thousands of headlines using these criteria, validating each other’s work to identify a large set of clickbait headlines.

From there, we built a system that looks at the set of clickbait headlines to determine what phrases are commonly used in clickbait headlines that are not used in other headlines. This is similar to how many email spam filters work. Our system identifies posts that are clickbait and which web domains and Pages these posts come from. Links posted from or shared from Pages or domains that consistently post clickbait headlines will appear lower in News Feed.