New method for detecting propaganda: Bochum researchers in focus!

New method for detecting propaganda: Bochum researchers in focus!

Disinformation campaigns are a growing problem that not only affects traditional media, but is also widespread in messenger services such as WhatsApp, Telegram and Signal. On August 13, 2025, a team of researchers from the Ruhr University Bochum and the University of Lausanne presented a newly developed method for detecting propaganda accounts on Telegram at the USENIX Security Symposium in Seattle. This method has the potential to effectively support the fight against disinformation.

The presentation led by Prof. Dr. Rebekah Overdorf stood out because the team of authors was awarded the prestigious Distinguished Paper Award. The methodology is not only faster and cheaper than previous human moderation, but also significantly exceeds it in terms of efficiency. By analyzing 13.7 million comments from 13 Telegram channels with a political or news focus, the researchers identified that 1.8 percent of these comments could be classified as propaganda, mainly from a pro-Russian network.

Efficiency of the new detection mechanism

The newly developed detection mechanism was able to detect propaganda with a hit rate of 97.6 percent, which is 11.6 percent better than the average efficiency of human moderation. This varies greatly, as some channels only have a detection rate of 20 percent, while others can reach over 95 percent. These differences illustrate how ineffective human moderation can be in combating disinformation.

In the course of this moderation, the high level of manual effort is also a problem. Content moderators are often exposed to psychological stress, especially when they have to deal with problematic content around the clock. The new algorithm could help reduce this effort and improve the working conditions of the moderators.

The context of disinformation

Combating disinformation is particularly relevant in times of crisis, where fake news and conspiracy narratives are spreading rapidly on social media. Experts believe that society is no more susceptible to these narratives than it used to be, but the presence of such content has increased significantly. This is often done with the strategic aim of stoking negative feelings such as fear or anger, which underlines the relevance of the newly developed methods for identifying propaganda.

According to a survey from the first quarter of 2023, 48 percent of Internet users in Germany said they had come across false or unreliable information. Disinformation has become a strategic tool used specifically to manipulate opinions, particularly in election campaigns. Despite these challenges, communication scientists emphasize that there are no acute threats to democracy from disinformation.

Every user can take active action against disinformation by questioning it critically, comparing sources and checking the senders of information. The European institutions have created clear regulations to make platforms such as Facebook, YouTube and TikTok responsible for dealing with the issue of disinformation. The Digital Services Act requires these platforms to assess risks to their users and take effective measures against disinformation.

The developments at the Ruhr University Bochum could be a decisive step towards a better information landscape in which disinformation can not only be discovered more quickly but also combated more effectively. A glimmer of hope in a time characterized by uncertainty and information overload.