My research focuses on the transformations of political journalism in France over the last decades, with a special focus on the changing framing of politics-related events in French print media. By combining machine learning (especially natural language processing) and econometric methods, I aim to identify the driving forces accounting for changes in journalistic norms.
I also have a methodological interest in computational social sciences. Based on examples from my empirical research, I attempt to illustrate how machine learning can provide new insights into some classic research questions in social sciences.
Check the package AugmentedSocialScientist to easily train BERT-like models!