Communications of the ACM
AI System Could Help Counter the Spread of Disinformation
The Reconnaissance of Influence Operations (RIO) research team at Lincoln Laboratory mapped a foreign Twitter propaganda network that was active during the 2017 French presidential election.
Credit: Lincoln Laboratory research team
The Reconnaissance of Influence Operations (RIO) program created by researchers at the Massachusetts Institute of Technology was intended to identify disinformation narratives and the individuals spreading them via social media.
In a study of the 2017 French elections, the researchers collected real-time social media data during the 30 days leading up for the election, compiling 28 million Twitter posts from 1 million accounts.
They found that the RIO system identified disinformation accounts with an accuracy rate of 96%.
RIO produces a comprehensive view of where and how disinformation narratives are spreading using a combination of analytic techniques.
Among other things, RIO uses machine learning to classify disinformation accounts based on factors like interactions with foreign media and the language they use.
It also can detect accounts operated by both bots and humans and quantify their impact on the social network.
From MIT News
View Full Article
Abstracts Copyright © 2021 SmithBucklin, Washington, DC, USA
No entries found