Tuesday, 02 January 2024 12:17 GMT

Swiss Develop Algorithm To Filter Russian Telegram Propaganda


(MENAFN- Swissinfo) Swiss researchers have developed an algorithm to automatically detect propaganda on the messaging service Telegram. This could help channel moderators filter out unwanted messages automatically. This content was published on August 21, 2025 - 09:15 3 minutes SRF

+Get the most important news from Switzerland in your inbox

“In some countries, such as Russia and Ukraine, Telegram is one of the most important information platforms. Content moderation is carried out by the individuals who lead a group or channel. There is no independent fact-checking body,” said the Swiss Federal Institute for Technology Lausanne EPFL.

+ How Switzerland is caught up in the Russian propaganda machine

Therefore, moderators set their own rules and must often manually remove unwanted content, despite having software at their disposal that is often based on very simple criteria, such as the detection of obscene expressions, according to EPFL.

The new algorithm, presented in a recent study, was developed by researchers at EPFL and the Max Planck Institute in collaboration with Ruhr University Bochum. It makes it possible to identify accounts that systematically spread propaganda on Telegram.

Repetitious wording

“One of the most striking characteristics of propaganda accounts is that they post messages with the same wording in different places, sometimes on different channels. While regular accounts post unique messages, propaganda accounts form large networks that repeatedly disseminate the same content,” EPFL's Carmela Troncoso told Swiss public broadcaster RTS.

The researchers used these repetitions to program an automated detection mechanism.

A total of 13.7 million comments from 13 Telegram channels were analyzed, focusing on politics and current events. Approximately 1.8% of the comments were found to be propaganda. The majority came from a pro-Russian network, while a smaller portion was pro-Ukrainian.

The algorithm“was able to detect propaganda based on a single comment with a success rate of 97.6%, which is 11.6% better than a human,” writes EPFL.

The research team now hopes that their tool can help ease the work of moderators, as they aren't available 24/7. Constant exposure to propaganda can also have negative effects on mental health.

More More Foreign Affairs Debunking Russian propaganda on Switzerland

This content was published on Sep 18, 2024 Russian authorities and official media claim that Swiss sanctions on Russia and Russian assets violate Switzerland's neutrality. SWI swissinfo fact checks these claims.

Read more: Debunking Russian propaganda on Switzerlan

MENAFN21082025000210011054ID1109956988

Legal Disclaimer:
MENAFN provides the information “as is” without warranty of any kind. We do not accept any responsibility or liability for the accuracy, content, images, videos, licenses, completeness, legality, or reliability of the information contained in this article. If you have any complaints or copyright issues related to this article, kindly contact the provider above.

Search