Antoine Lemor

This website is multilingual (FR/EN); you can change the language using the menu.

I am a Specialized Scientific Advisor at the Institut national de santé publique du Québec (INSPQ), where I contribute to the work of the Direction du développement des individus et des communautés (DDIC). I hold a Ph.D. in Political Science from Université de Montréal (Rector’s Honour List, Departmental Award for Best Dissertation). My postdoctoral research, funded by the FRQSC and hosted by the CIRST and the RFICS under the supervision of Professor François Claveau, examined the conditions under which scientific knowledge and technology gain influence over the policy process and reconfigure its dynamics. My research remains organised around these questions: evidence-informed policymaking, expert credibility, science-policy interfaces, and the impact of AI on the policy process.

I develop rigorous and reproducible computational methods, grounded in artificial intelligence, for the social sciences and public health, with the aim of making large-scale data analysis (text, images, audio, video) auditable and replicable through open-source tools and open research infrastructures. I lead the machine learning and technical development of Canadian Climate Framing (CCF), a live pan-Canadian observatory of climate media coverage holding more than 285,000 articles from 22 dailies, archived as they appear. Using methods grounded in artificial intelligence, we collect climate news every day and extract events and information from it with scientifically validated models published in Scientific Data (Nature Portfolio). With Tristan Boursier, I also co-lead YouPol, an observatory of political YouTube and TikTok that uses our own models to archive, transcribe and analyse more than 70,000 videos and 13 million comments from more than 100 channels in France, Quebec and the United States, including what the platforms remove. Like CCF, the platform is autonomous and built on local models. The database is accessible through an explorer or an API. I also created LLM Tool, an annotation tool combining locally run LLMs with BERT classifiers (video presentation, inference API).

My work has appeared in Governance, Policy & Politics, Policy and Society, Science and Public Policy, Scientific Data (Nature Portfolio) and the Revue française de science politique, as well as in edited volumes published by Edward Elgar and the Presses de l’Université de Montréal.

As part of the development of these methods, I hold research contracts with ENAP (École nationale d’administration publique), where I work on digital sovereignty and transnational data flows with Professor Guillaume Beaumier; Université de Montréal, where I develop methods for a project on the influence of emotions on the use of scientific evidence in parliamentary debates with Professor Éric Montpetit; and the Centre d’analyse des politiques publiques (CAPP, Université Laval), where I developed the machine learning models used by the vitrinedemocratique.com project. I teach political science and public policy analysis at UQAM and Université de Montréal, and I also deliver training on the use of computational methods and artificial intelligence for research.