Thesis of Maxime Le Duc

Co-construction and active learning applied to decision-making support


This work takes place thanks to a full cooperation between GRC Contact ( and the french laboratory LIRIS. The project is highly related to research in the artificial intelligence field, and aim specifically at developing an autonomous multi-agent system which would be able to learn in an unknown and changing world (human experts being part of this world), thanks to repeated interactions with it. By exploiting the learning capacities of this system, our main goal is to develop a decision-making support which will be integrated with a CRM (Customer Relationship Managment) and therefore be able to detect opportunities based on static and dynamic datas reinforced by human expertise.

Advisor: Frédéric Armetta
Coadvisor: Haytham Elghazel