University Grenoble Alpes, 2nd year Master of science in informatics (MoSIG), specialty Artificial intelligence and the web, Semantic web course

Semantics of Distributed Knowledge

Lecturers
Jérôme Euzenat (Jerome : Euzenat # inria : fr)
Language
English
Official web site
GBX9MO25

Teams

Objective

This is part of the "Semantic web: from XML to OWL" course (to be renamed: "Fundamentals of Data Processing and Distributed Knowledge"). Its starts from the observation that modern computing increasingly takes advantage of large amounts of distributed data and knowledge. This is grounded on theoretical principles borrowing to several fields of computer science such as programming languages, data bases, structured documentation, logic and artificial intelligence. The goal of this course is to present some of them, the problems that they solve and those that they uncover. The course considers three perspectives on data and knowledge: interpretation (what they mean), analysis (what they reveal) and processing (how can they be traversed efficiently and transformed safely).

The "Semantics of Distributed Knowledge" part offers a semantic perspective on distributed knowledge. Distributed knowledge may come from data sources using different ontologies on the semantic web, autonomous software agents learning knowledge or social robots interacting with different interlocutors. The course adopts a synthetic view on these and revolves around the question: how to approach the knowledge of a multiplicity of agents. It touches disciplines such as: knowledge representation, query evaluation, multi agent systems, logics and reasoning. The constant preoccupation of this course is how such disciplines can help and take advantage of distributed knowledge according to its semantics.

The course is organised around 5 topics. It first presents principles of the semantics of knowledge representation (RDF, OWL). Ontology alignments are then introduced to reduce the heterogeneity between distributed knowledge and their exploitation for answering federated queries is presented. A practical way for cooperating agents to evolve their knowledge is cultural knowledge evolution that is then illustrated. Finally, the course defines dynamic epistemic logics as a way to model the communication of knowledge and beliefs.

This opens perspectives to other types of activities such as belief revision and human-machine cooperation.

Place and time

Lectures are on Wednesday from 9h45 to 12h45.

Planning (2020–2021)

This can be consulted on the official timetable

DateTitleRoomLecturer
30/09/2020Knowledge, web, agents, etc.C006-Amphi (V)JE
21/10/2020Ontology networksH000-Amphi H (V)JE
04/11/2020Distributed query evaluationH000-Amphi H (V)JE
18/11/2020Social and cultural knowledge evolutionH000-Amphi H (V)JE
25/11/2020Logics of knowledgeH000-Amphi H (V)JE

Sillabus

The course is organised around 5 different topics corresponding to as many sessions.

Knowledge, web, agents, etc.

In which we start representing knowledge on computer, first independently, then together. Assessing meaning to this representation requires a semantics. This sometimes reveal that our knowledge is contradictory, redundant or misaligned.

  1. Distributed knowledge representation
    1. Motivations
    2. Occurence in the semantic web and modern AI
    3. Objectives of the course
  2. Expressing data with RDF: syntax and semantics
    1. RDF Graphs
    2. RDF Semantics
    3. Simple entailment
  3. Modelling knowledge with a simple ontology language
    1. A reduced ontology language
    2. Semantics
    3. Consistency, data and formula entailment

Indicative bibliography:

  • The lecture notes of the previous course, especially Part I.
  • Notes of the KRR course
  • Grigoris Antoniou, Frank van Harmelen, A semantic web primer, The MIT press, 2004 (rev. 2008)
  • Pascal Hitzler, Markus Krötzsch, Sebastian Rudolph, Foundations of semantic web technologies, Chapman & Hall/CRC, 2009
  • Networks of ontologies

    In which we lay bridges (alignments) between theories to address misunderstanding, ending with a network of related ontologies. We provide different semantics for such networks allowing to understand the benefit of alignments for interoperation. We can interpret the network as a whole or in the neighborhood of a particular agent.

    1. Connecting ontologies with alignments
      1. Ontology alignments
      2. Networks of ontologies
    2. Semantics of ontology alignments
      1. ADVANCED (not covered): Three alignment semantics
      2. The reduced alignment semantics
      3. Correspondence and formula entailment
      4. Alignment entailment entailment
    3. Semantics of networks of ontologies
      1. Semantics
      2. Overall consequences of alignment semantics
      3. Global and local consistency
    4. AVANCED (not covered): belief revision of networks of ontologies
      1. AGM postulates for belief revision
      2. Adaptation to description logics
      3. Semantic description of revision
      4. Problem specific to networks of ontologies

    Indicative bibliography:

  • The lecture notes of the previous course, especially Part III.8–10
  • Jérôme Euzenat, Pavel Shvaiko, Ontology matching, Springer Verlag, Heidelberg (DE), 2007; 2nde edition, 2013
  • Jérôme Euzenat, Revision in networks of ontologies, Artificial intelligence 228:195-216, 2015
  • Distributed queries

    In which we cross these bridges, exploiting networks of ontologies and linked data, in order to carry back answers to queries. The relationship between query evaluation and semantics becomes clear.

    1. Queries and entailment
      1. Graph query language
      2. Query semantics
    2. Querying modulo ontologies
    3. Queries over several data sources
    4. The different distributed query semantics
    5. Dealing with heterogeneous sources

    Indicative bibliography:

  • The lecture notes of the previous course, especially Part II.4–5 and III.12
  • Serge Abiteboul, Ioana Manolescu, Philippe Rigaux, Marie-Christine Rousset, Pierre Senellart, Data Integration, Chapter 9 of Web Data Management, Cambridge university press, Cambridge (UK), 2011
  • Philippe Adjiman, Philippe Chatalic, Francois Goasdoué, Marie-Christine Rousset, Laurent Simon, Distributed Reasoning in a Peer-to-Peer Setting: Application to the Semantic Web, Journal of Artificial Intelligence Research (JAIR) Volume 25. 2006.
  • Social and cultural knowledge evolution

    In which relying on existing bridges between knowledge representations is not sufficient: common knowledge is grounded on common experience. We investigate how dynamic techniques of cultural evolution can be compatible with knowledge representation semantics.

    1. Cultural knowledge evolution
      1. Motivation
      2. History
      3. Replicator-interactor
    2. Methodology
    3. The cultural alignment repair game
      1. Setting
      2. Adaptation operators and modalities
      3. Measured properties
    4. First results
      1. Convergence
      2. Operator comparison
      3. Comparison with logical repair
      4. Non scalability
    5. Extensions
      1. Relaxation
      2. Expansion
      3. Strengthening
      4. Starting empty
      5. Population variants

    Indicative bibliography:

  • Luc Steels (ed.), Experiments in cultural language evolution, John Benjamins, Amsterdam (NL), 2012
  • Alex Mesoudi, Andrew Whiten, Kevin Laland, Towards a unfied science of cultural evolution, Behavioral and brain sciences 29(4):329–383, 2006
  • Jérôme Euzenat, Interaction-based ontology alignment repair with expansion and relaxation, in: Proc. 26th International Joint Conference on Artificial Intelligence (IJCAI), Melbourne (VIC AU), 2017
  • Logics of knowledge

    In which these attempts are recast in decades of trying to capture knowledge into logic. Knowledge is too wild to be captured, but logic provides a solid tutor to grow along.

    1. Introduction to the semantics of modal logics
      1. Syntax
      2. Semantic structures and "pointed" models
      3. Ontology encoding
      4. Axiom schemata
    2. Epistemic-doxastic logic
      1. Syntax and axiomatics
      2. Baltag-Smets semantics
    3. Multi-agent epistemic logic
      1. Syntax and semantics
      2. Distributed and common knowledge
    4. Communication as action: dynamic epistemic logic
      1. Announcement and model transformation
    5. Example: modelling the alignment repair game
      1. Encoding ontologies as knowledge and alignments as belief
      2. Adaptation operators as announcements
      3. Logical properties of the alignment repair game

    Indicative bibliography:

  • Ronald Fagin, Joseph Halpern, Yoran Moses, Moshe Vardi, Reasoning about knowledge, The MIT press, Cambridge (MA US), 1995
  • Hans van Ditmarsch, Wiebe van der Hoek, Barteld Kool, Dynamic epistemic logic, Springer, Heidelberg (DE), 2018
  • Previous exams

    You may have a look at those of the previous sessions, however, the topic will now be largely different.