MOSIG Master 2ND YEAR Research
YEAR 2023–2024
Agents can adapt their knowledge using different adaptation operators and modalities. These operators may be considered as knowledge, so they also may be adapted and selected.
Cultural evolution is the application of evolution theory to culture [Messoudi 2006]. It may be addressed through multi-agent simulations [Steels 2012]. Experimental cultural evolution provides a population of agents with interaction games that are played randomly. In reaction to the outcome of such games, agents adapt their knowledge. It is possible to test hypotheses by precisely crafting the rules used by agents in games and observing the consequences.
Our ambition is to understand and develop general mechanisms by which a society evolves its knowledge. For that purpose, this approach has been adapted to the evolution of the way agents represent knowledge [Anslow & Rovatsos, 2015; Chocron & Schorlemmer, 2016; Euzenat, 2017; Bourahla et al., 2021]. In particular, we considered agents correcting their ontologies representing their environment in order to cooperate with other agents. We showed that agent knowledge is able to converge towards successful communication and improves the objective knowledge correctness [Bourahla et al., 2021].
Adaptation relies on adaptation operators using the outcome of failed tentative communication in order to behave differently in future interaction. They modify class descriptions in the ontology in order to better discriminate objects and make different decisions about them. We have proposed different adaptation operators. Although they are used by agents to select knowledge, they can also be considered knowledge, hence selected.
The goal of this research topic is to consider agents able to use any of these operators. Then experiments may be designed so that agents not only select target knowledge, but as well select the operators for selecting knowledge among a library of available methods. The objective qualities of various operators may be evaluated by the experimenter, but they can also be approximated by the agents.
They may do this judging by various criterion: speed, success, coverage. For instance, agents may try different operators and record whether their action lead to always successful interaction or if the classes need further refining that could have been avoided. We want to study the impact of this process on similar long term criterion applied to target knowledge.
An alternative is to have experiments using such agents and having the environment selecting those with the fittest choices (by harming or killing the others). It would be interesting to study the differences between the two approaches.
References:
[Anslow & Rovatsos, 2015] Michael Anslow, Michael Rovatsos, Aligning experientially grounded ontologies using language games, Proc. 4th international workshop on graph structure for knowledge representation, Buenos Aires (AR), pp15-31, 2015 [DOI:10.1007/978-3-319-28702-7_2]
[Bourahla et al., 2021] Yasser Bourahla, Manuel Atencia, Jérôme Euzenat, Knowledge improvement and diversity under interaction-driven adaptation of learned ontologies, Proc. 20th ACM international conference on Autonomous Agents and Multi-Agent Systems (AAMAS), London (UK), pp242-250, 2021 https://moex.inria.fr/files/papers/bourahla2021a.pdf
[Chocron & Schorlemmer, 2016] Paula Chocron, Marco Schorlemmer, Attuning ontology alignments to semantically heterogeneous multi-agent interactions, Proc. 22nd European Conference on Artificial Intelligence, Der Haague (NL), pp871-879, 2016 [DOI:10.3233/978-1-61499-672-9-871]
[Euzenat, 2017] Jérôme Euzenat, Communication-driven ontology alignment repair and expansion, in: Proc. 26th International joint conference on artificial intelligence (IJCAI), Melbourne (AU), pp185-191, 2017 https://moex.inria.fr/files/papers/euzenat2017a.pdf
[Mesoudi 2006] Alex Mesoudi, Andrew Whiten, Kevin Laland, Towards a unfied science of cultural evolution, Behavioral and brain sciences 29(4):329–383, 2006 http://alexmesoudi.com/s/Mesoudi_Whiten_Laland_BBS_2006.pdf
[Steels, 2012] Luc Steels (ed.), Experiments in cultural language evolution, John Benjamins, Amsterdam (NL), 2012
Links:
Reference number: Proposal n°3067
Master profile: M2R MOSIG, M2R MSIAM, M2R Informatics.
Advisor: Jérôme Euzenat (Jerome:Euzenat#inria:fr).
Team: The work will be carried out in the mOeX team common to INRIA & Université Grenoble Alpes. mOeX is dedicated to study knowledge evolution through adaptation. It gather permanent researchers from the Exmo team which has taken an active part these past 15 years in the development of the semantic web and more specifically ontology matching.
Laboratory: LIG.
Procedure: Contact us and provide vitæ and possibly motivation letter and references.