• Priority Area Supporting Research into Media, Disinformation and Information Literacy Across Europe
• Year 2023, 2024, 2025
• Country United Kingdom
• Project Status Ongoing

Online disinformation is a major challenge in our society, with potential to cause economical, social and public health harms. Disinformation narratives can evolve and reemerge over time and also be disseminated in multiple languages in multiple countries, often being adapted from one language into others. This hinders the work of fact-checkers, journalists and other stakeholders, working hard to debunk false and misleading information. Therefore, automatic multilingual and user-centric approaches to processing and analysing online disinformation narratives are urgently needed.

ExU (AI Models for Examining Multilingual Disinformation Narratives and Understanding their Spread) focuses on developing AI-based models for multilingual disinformation analysis, addressing the tasks of rumour stance classification and claim retrieval (all highly relevant to the process of fact-checking). ExU’s project partners, USFD and KInIT will develop cutting-edge AI tools, while also bringing together relevant contacts with journalists and fact-checkers through their participation in EDMO hubs and other HORIZON EU projects. Besides English, ExU will work with a set of 20+ languages, providing evaluation frameworks for at least seven: Portuguese, Spanish, French, Hindi, Polish, Slovak and Czech.

The novelty of the ExU project is two-folded: (1) it is the first effort on multilingual disinformation analysis considering various language families, relevant to the European scenario, and (2) it proposes a user-centric evaluation built on the top of state-of-the-art generative language models (like ChatGPT), focusing on providing the end users with the relevant information for understanding the outputs of ExU Models. This project is highly relevant to the EU and EDMO hubs, since it addresses languages spoken in multiple EU countries and also to other EU projects addressing disinformation analysis, since its results can be adapted to specific projects’ domains.