MOSIG Master 2ND YEAR Research
YEAR 2017/2018

Ontology and alignment coevolution through agent communication

Master topic / Sujet de master recherche

Ontologies are formal representations of entities that can be found in the world. As the world and our standpoint on it change, ontologies cannot remain static. We aim at evolving ontologies while they are used by agents for communicating about the world they inhabit. This may be achieved by exchanging pieces of ontologies or by adapting those ontologies which prevent efficient communication.

This may be approached through the framework of cultural evolution. 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.

The cultural language evolution approach [Steels, 2012] has been applied to the evolution of the way agents represent knowledge [Euzenat, 2014; Anslow & Rovatsos, 2015; Chocron & Schorlemmer, 2016]. We have applied this approach to ontology alignment repair, i.e., the improvement of incorrect alignments [Euzenat, 2014; 2017a; b]. We showed that cultural repair is able to converge towards successful communication through improving the objective correctness of alignments.

The goal of this master topic is to pursue this line of research through focussing on altering ontologies —as opposed to alignments— following errors in communications between agents. This involves considering how and when agents may adopt the ontologies of each other or discard parts of their ontologies and when they may do so. More sophisticated multi-agent protocols may be required. Explication protocols [van Diggelen & al., 2007] allow agents to provide examples when they realise that the other agent will not understand them; gossiping protocols may generalise our experiments to whole populations of connected agents [Aberer & al., 2003].

For instance, agents may fail to understand a concept (food) used by another agent and request examples of these concepts (apple, potatoes). Then with respect to the position of the example in their own ontology (vegetables) they may be able to interpret it. However, this understanding is to be later revised if provided with other examples (prawn) which do not fit in the same class: indeed, food is not a subclass of vegetables. Agents interacting with several peers may feel the need to reengineer their ontologies depending on its capability to communicate (some classes may be useless, others may be required).

Expected contributions are firstly the design of an experimental scenario (based on what we already have) and ontology adaptation operators in such a situation. They will be used to investigate, experimentally and/or theoretically, properties of such operators, e.g., convergence to a stable state, suppression of communication errors.

This work is part of an ambitious program towards what we call cultural knowledge evolution and may prepare to a PhD.

References:

[Aberer & al., 2003] Karl Aberer, Philippe Cudré-Mauroux, Manfred Hauswirth, Start making sense: The Chatty Web approach for global semantic agreements, Journal of web semantics 1(1):89-114, 2003
[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]
[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 & Shvaiko, 2013] Jérôme Euzenat, Pavel Shvaiko, Ontology matching, 2nd edition, Springer-Verlag, Heildelberg (DE), 2013
[Euzenat, 2014] Jérôme Euzenat, First experiments in cultural alignment repair (extended version), in: Proc. 3rd ESWC workshop on Debugging ontologies and ontology mappings (WoDOOM), Hersounisos (GR), LNCS 8798:115-130, 2014 ftp://ftp.inrialpes.fr/pub/exmo/publications/euzenat2014c.pdf
[Euzenat, 2017a] Jérôme Euzenat, Communication-driven ontology alignment repair and expansion, in: Proc. 26th International joint conference on artificial intelligence (IJCAI), Melbourne (AU), pp185-191, 2017 ftp://ftp.inrialpes.fr/pub/moex/papers/euzenat2017a.pdf
[Euzenat, 2017b] Jérôme Euzenat, Crafting ontology alignments from scratch through agent communication, in: Proc. 20th International conference on principles and practive of multiagent systems (PRIMA), Nice (FR), pp245-262, 2017 ftp://ftp.inrialpes.fr/pub/moex/papers/euzenat2017b.pdf
[Steels, 2012] Luc Steels (ed.), Experiments in cultural language evolution, John Benjamins, Amsterdam (NL), 2012
[van Diggelen & al., 2007] Jurriaan van Diggelen, Robbert-Jan Beun, Frank Dignum, Rogier van Eijk, John-Jules Meyer, Ontology negotiation in heterogeneous multi-agent systems: The ANEMONE system, Applied ontology 2(3-4):267-303, 2007

Links:


Reference number: Proposal n°2242

Master profile: M2R MOSIG, Artificial intelligence and the web profile.

Advisor: Jérôme Euzenat (Jerome:Euzenat#inria:fr) and Manuel Atencia (Manuel:Atencia#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, motivation letter and references.