Digital Planet

By BBC World Service

Tracking the trolls

🔊 Play episode (42 min)

Direct episode link

💬 Share episode

Published July 28, 2020 3:00pm

How can we distinguish the online posts written by real people from those coming out of professional bot-farms intent on influencing elections? New research from Princeton University in America uses machine learning to identify malicious online trolls, even before they’ve sent a single tweet. Lead author Meysam Alizadeh explains the power of this work to protect voters in future elections.

Gesture-controlled robotsRobots can now be controlled by a simple wave of your arm. Professor Daniela Rus from MIT explains how new research has simplified robot controls by using human movement rather than complicated systems of bu...

Return to podcast