Deborah Dore
PhD Student in the Marianne TEAM.
E416 - Bâtiment E
Université Côte d'Azur
06410 Biot
My background is rooted in Artificial Intelligence. During my Bachelor’s, I studied core topics in Computer Science such as Databases, Algorithms, and Programming. In my Master’s, I specialized in AI through courses in Machine Learning, Deep Learning, and Natural Language Processing, which naturally led me to pursue a Ph.D.
Currently, I focus on Argument Mining, a subfield of NLP concerned with automatically identifying arguments in text and classifying their structure—claims, premises, and their supporting or attacking relationships. This technology can help expose logical fallacies and propaganda, fostering more informed public discourse.
My research applies Argument Mining to political debates, where speakers build complex argumentative structures. Mapping these interactions can support voters in making better-informed decisions.
news
| May 28, 2025 | 📢 My Master’s thesis work, “Leveraging Graph Structural Knowledge to Improve Argument Relation Prediction in Political Debates” (Deborah Dore, Stefano Faralli, Serena Villata), has been accepted at the Argument Mining Workshop at ACL 2025! In this research, we investigate how integrating Knowledge Graph Embedding Models with Large Language Models can advance the state of the art in predicting argumentative relations within US political debates. |
|---|---|
| May 23, 2025 | 📢 Our paper “DISPUTool 3.0: Fallacy Detection and Repairing in Argumentative Political Debates” (Pierpaolo Goffredo, Deborah Dore, Elena Cabrio, Serena Villata) has been accepted to ACL 2025’s Demo Track! DISPUTool 3.0 is a web-based tool that detects and repairs fallacious arguments in political debates. Built on a collection of U.S. presidential debates, from 1960 to 2020, DISPUTool goes beyond detection by suggesting revised, fallacy-free versions of flawed arguments. |
| Apr 25, 2025 | 📢 Our paper “Repairing Fallacious Argumentation in Political Debates” (Pierpaolo Goffredo, Deborah Dore, Elena Cabrio, Serena Villata) has been accepted to ECA 2025! In this work, we explore how to identify and correct fallacies in political debates using computational approaches — a step toward improving the quality of public debate. |
| Jan 01, 2025 | 🚀 A new chapter begins! We’ve officially branched out from WIMMICS to form MARIANNE — a new research group dedicated to argument mining, political debates, and misinformation. |
| Nov 01, 2024 | 🎓 I started my PhD at the Université Côte d’Azur! My work will focus on Argument Mining on Political Debates. This research is supported by the 3IA Côte d’Azur! |