Note
The repositories are private. To request private access, please contact rayane.ghilene@ensea.fr.
The objective of Arenas is to characterize, measure, and understand the role of extremist narratives in discourses that have an impact not only on political and social spheres but importantly on the stakeholders themselves. Leading an innovative and ambitious research program, ARENAS will significantly contribute to filling the gap in contemporary research, make recommendations to policymakers, media, lawyers, social inclusion professionals, and educational institutions, and propose solutions for countering extreme narratives for developing more inclusive and respectful European societies.
Zero-Shot SUD Classification
An entailment framework using NLI models for unsupervised discourse classification.
📄 arXiv preprint
Machine Learning is heading to the Socially Unacceptable Discourse analysis
From Shallow Learning to Large Language Models to the rescue, where do we stand? (Benchmark)
Artifact-Guided Pretraining
Upcoming release (code and paper available after publication acceptance).
Unsupervised Graph Structure Learning
An unsupervised method for understanding hidden structures in discourse graphs.
Michele LINARDI: Assistant Professor of Computer Science at the ETIS Lab and the IUT (University Institutes of Technology) of CY Cergy Paris Université.
Mohamed Rayane GHILENE: NLP Research Engineer at the Multidisciplinary Research Centre for the Human and Social Sciences CY Cergy Paris Université.
Dimitra NIAOURI: PhD Candidate at CY Cergy Paris Université within the ETIS and AGORA labs.
Julien Longhi: Professor of Discourse Analysis and Digital Humanities. Coordinator of the ARENAS project and work package leader.
If you have any questions, please don't hesitate to contact rayane.ghilene@ensea.fr and dimitra.niaouri@cyu.fr.
These resources can be used for research purposes using the Creative Commons Attribution 4.0 License. Please cite the corresponding publication when publishing.