Network based debates currently have an enormous influence on the political public. Russian trolls who try to influence the outcome of the US elections, so-called “social bots” that secretly spread opinion changing propaganda, and a much more agitated speech even including hate posts; “the internet is broken” is a commonly heard phrase. But is it? What are the consequences of the current developments? Are bots and their algorithms going to decide the German elections of 2017? Which technical or societal remedies do we have? What are the dangers and chances of introducing more control into internet debates? On November 9, ERCIS and IfK (communication sciences institute) researchers have organized an open discussion even that was announced on the university level. The debate has been seeded by three presentations:
- Sebastian Köffer on the question “can analytics reduce network hatred”? He reported on the results of the cyberhate mining project and the methods developed therein for recognizing hate speech.
- Lena Frischlich reported on Hate and Propaganda on the internet from a communication science perspective, especially concerning possible societal effects of (automated) propaganda.
- Christian Grimme, Mike Preuß and Lena Adam provided a technically oriented view onto social bots. Via explaining a simple self-made Twitter bot named “John Potts” they explained how easy it is to create automated tweets.
The following, lively discussion was held in the fishbowl format and included also several participants from the audience. Cyberhate and propaganda are the main focus of two current research projects at the WWU Münster: The project PropStop (in collaboration with the communication sciences, the Technical University of Braunschweig and industry partners) deals with detection and the establishment of countermeasures against secret Propaganda attacks via online media. The project Cyberhate-Mining investigated internet hate speech during the refugee crisis of 2015.