Welcome to my... webpage?! I am a Ph.D. student in LAMSADE (CNRS UMR 7243) at Université Paris-Dauphine since November 2017. I am currently working on Abstract argumentation theory and agent-based modeling with my supervisors Juliette Rouchier and Gabriella Pigozzi . The title of my proposal goes: "Les Corses avaient-ils raison? La théorie de l’argumentation et la simulation agents comme aides à l’analyse d’un conflit environnemental et des dynamiques politiques à plusieurs échelles".
The current goal of my thesis is to find comprehensive complex systems to analyze social conflicts and produce recommendations on how to avoid or eventually solve them. In the thesis, I attempt to understand how deliberation, discussion, and aggregation rules affect collective decision-making. I try to find a satisfying ontology that links argumentation, opinion formation, and collective decision-making using agent-based models of opinion diffusion and abstract argumentation theory. The main idea behind the ontology is that opinions and preferences on alteranatives are not only determined by individual affect for one alternative or another but also on how the affect is justified. Preferences and opinions are not exogenous and are constructed during social interactions such as deliberation arenas or casual discussion with peers, which may lead to band-wagon effects or clivages. Well-justified affects can be very influential and lead to updates of preferences, opinions, and voting intentions .
On the side, I study non-monotonic logics and belief revision theories for agent-based modeling and peer-effect structures and their effect on collective decision-making in Condorcet jury frameworks.
Recent work on my Ph.D. thesis. The paper describes an opinion diffusion model in which well-justified arguments can result in an update in opinions. It shows that deliberation tends to bring about consensus in opinion depending on a wide set of behavioral and procedural parameters.
Draft of my Master's dissertation in Economics. It gives insight on the question of peer-effects in small networks in collective decision-making. The model I presented showed to what extent correlation among votes can be deterimental to voters.
Draft of a manuscript (in French) on the Double-Dividend hypothesis and fish resources. Using agent-based modeling techniques, the model yields that the introduction of taxes on rejected fish can keep fish stocks from depleting.
In this manuscript (in French), I present an agent-based model of innovation diffusion and homophily inspired from classic models of innovation diffusion. I show that homophily in the update of networks has little to no effect on the diffusion of an innovation. Market effects and the price function of the innovation dominate the diffusion.
Research note (in French) on the relationship between human capital and salaries in France. I took the "Generation 98" database and checked if some basic postulates in the the theory of human capital held. On average, more years of education and more professional experience imply higher salaries.
Using the "Generation 98" French database, I show that, on average, individuals with higher education get permanent jobs faster than individuals with less education. However, individuals with technical education get permanent jobs as fast as individuals with Master's degrees and faster than PhDs.
In this manuscript (in French) a collegue, Adama Ouedraogo, and I study the feeling of insecurity caused by immigration and terrorism in Germany. Our analyses of the SOEP 2013 survey show that worries about global terrorism are linked, for the most part, to age and gender. This validates the vulnerability hypothesis defended in the literature on crime. Additionally, we find that individual worries about global terrorism are also dependent on the level of education and income.
Old first draft manuscript on well-being and macroeconomic indicators of performance. It shows that an aggregated subjective life satisfaction statistic is negatively correlated with unemployment and positively correlated with consumption. I have not touched the manuscript since 2015 and has many mistakes, but it will hopefully be upgraded...