Representing Belief Revision through Default Theories

  • Authors:
  • João C. P. da Silva;Sheila R. M. Veloso

  • Affiliations:
  • -;-

  • Venue:
  • IBERAMIA-SBIA '00 Proceedings of the International Joint Conference, 7th Ibero-American Conference on AI: Advances in Artificial Intelligence
  • Year:
  • 2000

Quantified Score

Hi-index 0.00

Visualization

Abstract

We examine the representations of epistemic states and their revision processes (RP) through default theories (DTs). Using a sphere-based semantics for DTs we characterize a RP as a sequence of consistent expansions without using contractions. We show that DTs provide an unifying framework for revision in both semantical and syntactical-based approaches as well as in coherent and foundation approaches. Motivated by Nebel's work, we propose to extend the use of default theories in two directions: (i) using supernormal default theories to represent semantical classified revision processes; (ii) using general default theories and some variants of Reiter's default logics to represent foundation classified revision processes. We define revision operators, which can be viewed as a revision on the epistemic states, represented by a DT T, by a sentence α Basically, they are defined via some operations on the class of the extensions of a default theory T' obtained from T and α.