• HDR Defended on :
  • Nov 27, 2019 • salle des thèses, Faculté des sciences Jean Perrin

Abstract

Our research work is situated in the framework of artificial intelligence. It concerns the use of formal logics, as well as other symbolic formalisms that are related to them, to answer different problems, and this, around two main themes: data mining by declarative approaches and knowledge representation. Our work is thus transversal to the two axes of the Centre de Recherche en Informatique de Lens (CRIL), namely “Algorithms for inference and constraints” and “Knowledge representation and reasoning”. Indeed, our contributions are related to the modeling of different data mining problems in classical propositional logic, reasoning in the presence of inconsistency, argumentation theory and qualitative reasoning. Because of its preponderant place in all our works, our presentation is mainly devoted to the use of classical propositional logic as a modeling tool. The latter is articulated around three main angles of view. The first angle concerns the use of the consistency problem in propositional logic. The second angle is related to optimization problems derived from propositional logic. The third angle is related to the use of this logic in the presence of incoherence.

Rapporteurs

  • Bruno Crémilleux, Professor at the University of Caen Normandy
  • Frédéric Saubion, Professor at the University of Angers
  • Torsten Schaub, Professor at the University of Potsdam

Reviewers

  • Salima Benbernou, Professor at the University of Paris Descartes
  • Jean-François Condotta, Professor at the University of Artois
  • Jean-Marc Petit, Professor at the National Institute of Applied Sciences of Lyon
  • Souhila Kaci, Professor at the University of Montpellier
  • Lakhdar Sais, Professor at the University of Artois