Imagine scrolling through your social media feed and seeing a bunch of posts about politics. You might not know it, but some of those posts could be part of something called computational propaganda.

     Computational propaganda is when political actors use fancy tricks to influence what you see and think.

     This can include using fake accounts, or "bots" or multiple accounts under the control of one person.

     This type of thing can make it difficult to know what's real and what's not, especially with the rise of large language models and the overwhelming amount of information online. Large language model refers to a type of artificial intelligence that understands and generates human-language text. It is a program designed to learn from and make predictions based on data, so it could be responsible for creating fake news stories and spreading misinformation more widely and more convincingly.

     Computational propaganda is tricky, because it often looks like real news shared by real people. It mixes real users with fake accounts, making it hard to spot.

      Recently, scientists tried a new approach to spot political propaganda on social media. They looked at 16 datasets, which included tens of thousands to a few million tweets. Out of these, 10 datasets were about politics, their themes ranged from the political groups to movements and causes. The other 6 were about non-political topics, like natural disasters.

     What did they discover? They found specific statistical and network patterns that help tell the difference between political and straightforward information. By focusing on things like how often hashtags were used and how users mentioned each other, they could spot the difference between political propaganda and everyday chatter.

     The scientists even built a fancy graph connecting words and tweets to better understand the flow of information. The most interesting part is that this method doesn't care where the tweets come from. Whether it's from a tweetstorm in Japan or a political debate in Brazil, the method stays the same.

     This isn't just about spotting political chatter; it could be used to understand any kind of social media trend, be they movements or memes. It could be anything from a concert or about the latest music hits.

    However, let's be clear, this work is just the beginning and the tool is far away from becoming the ultimate detective. There are vast complexity and magnitude of computational propaganda techniques are growing and get more and more sophisticated every day. It also doesn't mean we should stop thinking for ourselves and stop asking questions such as:

    Is this really true? Who's behind this information? Why are they sharing this?

     We need to stay critical and ask questions, even if a friend shares it or if it looks like real news.

Unmasking Political Spin on Social Media for Teens

Actions