[ad_1]
The social network owned by Facebook said that it was now using a machine learning technology to "proactively detect intimidation in photos and their captions". Technology marked posts are sent to human Instagram moderators for review.
Bullies may have more trouble spreading hatred on Instagram in the future.
In an article published Tuesday on the blog, the social network owned by Facebook said that he was now using a machine learning technology to "proactively detect intimidation in photos and their legends" . Technology marked posts are sent to human Instagram moderators for review.
"This change will help us identify and eliminate significantly more bullying – and this is a crucial step because many people who experience or observe intimidation do not report it," wrote Instagram manager Adam Mosseri. "It will also help us protect the youngest members of our community because teens suffer higher online bullying rates than others."
In May, Instagram began to proactively remove intimidation comments from the feeds, profiles, and the Explore People section. Now, the same goes for comments on live video.
Instagram has also just introduced a new "kindness camera effect" created in collaboration with author, dancer and teen actor Maddie Ziegler to help spread positivity on the platform. Ziegler fans will automatically have access to this new camera effect.
When you select the selfie mode effect, "the hearts will fill the screen – and you will be encouraged to tag a friend you wish to support." This person will then be able to use the camera effect to spread the love.
Meanwhile, Business Insider has recently detailed another problem regarding Instagram: children's videos being exploited on its IGTV service. An investigation conducted by the press briefing revealed that IGTV had recommended "graphic and disturbing" content, including "sexually suggestive sequences of young girls and an explicit video of a mutilated penis". Instagram said that it had removed all offensive content discovered by Business Insider.
[ad_2]
Source link
