Misinformation on Twitter and Reddit can now be stopped in its tracks
It has been thought that artificial intelligence (AI) could help deal with the big amount of trolling and hate that is flooding social media, with many efforts being made towards that goal. A team from Princeton University has developed an AI-based system that can be used to detect foreign trolls trying to influence election results in real-time. In the past, platforms like Twitter and Reddit have been used as a means to spread misinformation to people and alter voters perceptions.
The study was published in the journal Science Advances, and this AI-software was used to investigate related situations about misinformation campaigns that took place in China, Russia and Venezuela, with the campaigns being used as political tools against the US before and after the 2016 presidential elections. This new model was able to properly identify those posts that were part of a foreign influence campaign, including those that come from accounts that were never used before.
The process of developing this tool is still ongoing, and researchers hope that software engineers can get to the point that this design works as a real-time monitoring system for exposing influence American politics.
“It is not perfect, but it would force these actors to get more creative and possibly stop their efforts. You can only imagine how much better this could be if someone puts in the engineering efforts to optimise it,” said Jacob Shapiro, Professor at the Princeton School of Public and International Affairs. The data used to feed the AI system came from Twitter and Reddit because both of them contain a rich dataset of posts by users who are involved in politics as well as some data from the New York University’s Center for Social Media and Politics.
