SMILE Team
Multi-agent Systems, Interaction, Language, Evolution
SMILE’s activities are federated around the paradigm of Multi-Agent Systems (MAS). In various research scenarios, SMILE uses this paradigm to study, model and implement computer systems composed of interacting intelligent and autonomous entities. SMILE’s scientific expertise addresses issues related to the two modeling levels found in MAS: (1) the individual level, with models allowing the modeling and representation of behaviors (decision, learning, preferences, etc.) and (2) the collective level, with models of organizations and interactions allowing the efficient structuring of these systems. Depending on the case studies, the considered individuals can be either artificial agents (e.g. in robotics), or model of human agents (e.g. in preference representation), or both, as in the context of video games (serious or not) where the software components and the user can be understood as interacting agents participating in one and the same system.
Staff
Souhila Kaci, Professeur des universités, UM
Tiberiu Stratulat, Maître de conférences, UM
Abdelkader Gouaich, Maître de conférences, UM
Nicolas Serrurier Gourvès, Adjoint technique, UM
Fabien Michel, Maître de conférences, UM
Associates and Students
Noah Collinet, UM
Tiziano Maisonhaute, CIRAD
Hector Roussille, UM
Regular Co-workers
Stefano Alessandro Cerri, Invité longue durée Eméritat, UM
SMILE’s research is organized around the following axes:
- Games: this axis considers games as particular instances of MAS and studies their dynamics, especially to equip them with adaptation capacities allowing taking into account the characteristics of each user. Several fields of application are considered, including serious games for health, education and professional training.
- Reasoning and Representation: this axis studies models and mechanisms allowing the representation of artificial or natural reasoning and behaviour. Topics actually focused are reasoning taking into account preferences or beliefs, and models of argumentation between agents.
- Multi-Agent Modeling and Simulation: MAS are commonly used as a modeling paradigm for the study of complex systems. Our research covers both theoretical (formal and conceptual models) and practical (generic software tools) aspects, and is applied in various academic and industrial contexts.
- 2022 – Co-organisation of the first /Simulation and Artificial Intelligence Day, /under the aegis of the AFIA, with the theme /Simulation for AI and AI for simulation: https://afia.asso.fr/simulation-ia-2022/
- 2022 – Co-organisation of the 4th Journées Francophones de la Modélisation et de la Simulation (JFMS 2022): workshop of the DEVS-RED network: https://devs-network.org/jfms-2022/
- 2020 – Co-organisation of the 3rd Journées Francophones de la Modélisation et de la Simulation (JFMS 2020) – workshop of the DEVS-RED network: https://devs-network.org/posts/jfm-2020-ouverture/
- 2017 – Creation and maturation at the LIRMM of a startup: ADAGE project. The objective of the ADAGE project is the development of a professional publishing suite for adaptive serious games for training.
- 2016 – General chair of the 24th JFSMA : “Journées Francophones sur les Systèmes Multi-Agents”
- 2014 – Chair E4MAS (Agent Environments for Multi-agent Systems) – 10 Years Later @ AAMAS 2014
Title: Agents d’apprentissage par renforcement en profondeur révélant les incertitudes dans les systèmes à chaîne de blocs
PhD defendant: Hector Roussille
Defense date: 2024-12-06
Thesis director:
Fabien Michel
Title: Vers un Apprentissage Protéiforme : Accomoder les Changements de Signature des Agents Artificiels.
PhD defendant: Iago Bonnici
Defense date: 2021-03-29
Thesis director:
Fabien Michel
Title: Apprentissage épigénétique de comportements autonomes dans une société dagents
PhD defendant: François Suro
Defense date: 2020-09-17
Thesis directors:
Jacques Ferber,
Fabien Michel
Title: Aide à la construction et l’évaluation des preuves mathématiques déductives par les systèmes d’argumentation
PhD defendant: Nadira Boudjani
Defense date: 2018-12-05
Thesis director:
Souhila Kaci
Title: ViewointS : vers une émergence de connaissances collectives par élicitation de point de vue
PhD defendant: Guillaume Surroca
Defense date: 2017-06-30
Thesis director:
Stefano Alessandro Cerri
Title: Modélisation et implémentation de simulations multi-agents sur architectures massivement parallèles
PhD defendant: Emmanuel Hermellin
Defense date: 2016-11-18
Thesis directors:
Jacques Ferber,
Fabien Michel
Title: Vers l’enrichissement d’ontologies biomédicales françaises
PhD defendant: Juan Lossio
Defense date: 2015-11-09
Thesis directors:
Clement Jonquet,
Mathieu Roche
Title: Utiliser une culture et des valeurs pour coordonner flexiblement des agents
PhD defendant: Lois Vanhée
Defense date: 2015-09-22
Thesis director:
Jacques Ferber
Title: Adapter les communications des jeux dans le cloud.
PhD defendant: Richard Ewelle Ewelle
Defense date: 2015-08-28
Thesis director:
Stefano Alessandro Cerri