Kick-off Workshop: Building the POLAR Research Framework

This workshop brought together an interdisciplinary team from Paris, Antwerp, and Cyprus to establish a comprehensive research framework for identifying online controversy and polarization. By integrating Critical Discourse Analysis with advanced Natural Language Processing (NLP), the project aims to move beyond simple hate-speech detection to uncover covert polemical markers—such as irony, “just asking” strategies, and “us vs. them” framing—that trigger and escalate digital conflict. The discussions focused on how these linguistic nuances function across different “linguacultures” and how ideological frameworks govern the emotional escalation of online debates.

Moving forward, the project will execute a multi-phase plan starting with the creation of a comprehensive glossary of polemical strategies at linguistic, abstract, and argumentative levels. Leveraging datasets from platforms like X (formerly Twitter) and Reddit, the team will develop and test automated detection models involving both fine-tuned local transformers and Large Language Models (LLMs). The ultimate goal is to investigate tool-assisted decision-making and the potential for automatic counter-discourse generation to de-escalate polarized discussions and provide societal value.