Olivier Cailloux, Maitre de Conférences (meaning: researcher and teacher) in Computer Science.
Coordinates (at the Dauphine campus, 4th floor, see the map here).
Keywords of interest:
I am interested in studying legitimacy of grounds for recommending. Relatedly, I am interested in obtaining preference models in multicriteria contexts (Figueira, Greco, and Ehrgott 2005), preference agregation, and explanation of recommendations adapted to the user’s subjectivity.
In complex decision problems, preferences are not precisely determined in the head of the decision maker prior to the start of the modeling process: the preferences expressed crucially depend on the precise phrasing of questions, even among seemingly equivalent questions (Lichtenstein and Slovic 2006; Morton and Fasolo 2008). More generally, one can expect that by learning facts about her decision problem, or by viewing the decision problem under various angles, the individual will change her preference.
I am interested in studying the stability of changes of preferences. This is in contrast to the study of “shallow preferences” that is generally the center of attention and that takes no account of stability issues. Capturing preferences that are stable facing new information may be a way to ground normative models (models that give recommendations). This could be achieved using ideas borrowed from formal argumentation theory (Rahwan and Simari 2009). Argumentation theory studies the ways of selecting arguments among a set of arguments given an attack relation (a binary relation indicating how arguments conflict).
Voting theory is another interest of mine. I want to explore how we can help decision makers choose a suitable voting rule. We should develop strategies that permit to transmit the knowledge built by researchers in social choice to decision makers, without unduly orienting the decision, and while making the knowledge understandable to non-experts. We have extended preference elicitation techniques to the selection of a voting rule; and have worked on explaining the results of voting rules on concrete examples.