Date and Time: 25/06/26, 5pm CET
Speaker: Laura Alonso Alemany, Mathematics, Astronomy and Physics Faculty, Universidad Nacional de Córdoba (Argentina)
Presenter: Horacio Saggion, Pompeu Fabra University
Abstract:
Persuasion is one of the most elusive aspects when it comes to formalising natural language. Valuable formalisms have been proposed for analysing and generating discursive, argumentative, and narrative structures; by contrast, persuasion is often relegated to vague notions of "emotionality" or "group identity", or to inventories of rhetorical structures that typically prove difficult to apply computationally. However, the persuasive power we are currently observing in Large Language Models means we can no longer afford to overlook this dimension of text.
In this talk, we will discuss empirical results regarding the persuasive capabilities of language models and briefly describe several proposals for tackling rhetoric, argumentation, and persuasion. Finally, I will share some of my own experiences in generating counter-narratives and critical questions as alternative discursive forms to persuasion.
Bio: Laura Alonso Alemany holds a PhD in Computational Linguistics from the University of Barcelona and is a Professor of Computer Science at the Faculty of Mathematics, Astronomy, Physics, and Computing at the National University of Córdoba. She specialises in Natural Language Processing and AI Ethics. Her research interests focus on how empirical methods can help us understand natural language phenomena, as well as investigating the various real-world implications of errors within Artificial Intelligence systems.
Enlace a la charla: https://zoom.us/webinar/register/WN_ydR727a9SS2ieDM7EV0hmA
