Over the past years, we have had the opportunity to be part of the HYBRIDS project, a European Doctoral Network focused on understanding and addressing challenges such as disinformation, online manipulation, hate speech, and harmful digital communication through artificial intelligence and language technologies.
As an industry partner, our contribution has been to welcome PhD researchers into our company, providing them with the opportunity to connect their research with real-world challenges and explore how academic advances can translate into practical applications.
We are especially proud to have hosted five researchers whose work tackles some of the most relevant challenges in today’s digital landscape:
- Paloma Piot Pérez-Abadín (Universidade da Coruña)
Detecting harassment on social media beyond isolated messages
Paloma’s research focuses on identifying threatening and harmful online messages by considering not only individual posts, but also the wider conversational and interactional context. By incorporating previous messages, user behaviour, and multilingual analysis, her work aims to provide a more complete understanding of online harassment. - Martial Pastor (Radboud University)
Towards an automatic characterization of micro-level structures of discourse in short-form social media content
Martial’s research focuses on automating discourse analysis in short texts such as tweets and Reddit posts. His work addresses the challenges of identifying micro-level discourse units and modelling rhetorical relations between them, such as opposition, causality, and elaboration. The aim is to better understand how meaning and argumentation are structured in fragmented online communication. - Michele Joshua Maggini (CiTIUS - USC)
Detection of hyperpartisan political news
Michele’s research focuses on detecting and mitigating hyperpartisan political news by identifying highly biased and polarised content that can influence public perception of political issues and candidates. His work explores methods to promote information diversity by incorporating multiple perspectives on the same topic, as well as techniques for neutralising biased language through automatic paraphrasing into more objective and balanced formulations. - Katarina Laken (Fondazione Bruno Kessler)
Analysis of the role of disinformation to automatically identify hate speech against immigrants on multilingual social environments
Katarina’s research investigates strategies for detecting hate speech and disinformation, including prejudiced narratives linking immigrants to crime or disease, or reinforcing harmful stereotypes. Her work also explores how these phenomena appear in social media and how multilingual and cross-lingual approaches can help detect them more effectively. - Søren Fomsgaard (Université de Caen Normandie)
Detection of toxic bots in Twitter by analysing textual content
Søren’s research develops methods for identifying and characterising toxic and AI-generated content on Twitter, particularly in the context of political bots. His objectives include building classification systems to detect automated political actors, collecting and analysing campaign-related Twitter data, and extracting linguistic patterns such as topic bias, extremity of opinions, and stylistic markers. He also works on developing a taxonomy of bots based on their goals, behaviour, and discourse features.
Their work reflects the importance of combining AI, language technologies, and social analysis to better understand complex information environments and online harms.
Thank you, Paloma, Martial, Michele, Katarina and Søren, for being part of our team and for the knowledge and insights you have shared with us throughout this journey.
We look forward to joining the final HYBRIDS events in Trento next week and meeting the rest of the project community.
