A Class of df-consistencies for Qualitative Constraint Networks

Jean-Francois Condotta (1) , Christophe Lecoutre (1)

1 CRIL - Centre de Recherche en Informatique de Lens

In this paper, we introduce a new class of local consis- tencies, called °f-consistencies, for qualitative constraint networks. Each consistency of this class is based on weak composition (°) and a mapping f that provides a covering for each relation. We study the connections ex- isting between some properties of mappings f and the relative inference strength of °f-consistencies. The con- sistency obtained by the usual closure under weak com- position is shown to be the weakest element of the class, and new promising perspectives are shown to be opened by °f-consistencies stronger than weak composition. We also propose a generic algorithm that allows us to com- pute the closure of qualitative constraint networks un- der any "well-behaved" consistency of the class. The experimentation that we have conducted on qualitative constraint networks from the Interval Algebra shows the interest of these new local consistencies, in particular for the consistency problem.

Domaines: Intelligence artificielle [cs.AI]

Jean-Francois Condotta, Christophe Lecoutre. A Class of df-consistencies for Qualitative Constraint Networks. 12th International Conference on Principles of Knowledge Representation and Reasoning (KR'10), 2010, Toronto, Canada. pp.319-328. ⟨hal-00865375⟩