Extending Removed Sets Revision to partially preordered belief bases

Mariette Serayet (1) , Pierre Drap (1) , Odile Papini (1)

1 LSIS - Laboratoire des Sciences de l'Information et des Systèmes

Most of belief revision operations have been proposed for totally preordrered information. However, in case of partial ignorance, pieces of information are partially preordered and few effective approaches of revision have been proposed. The paper presents a new framework for revising partially preordered information, called Partially Preordered Removed Sets Revision (PPRSR). The notion of removed set, initially defined in the context of the revision of non ordered or totally preordered information is extended to partial preorders. The removed sets are efficiently computed thanks to a suitable encoding of the revision problem into logic programming with answer set semantics. This framework captures the possibilistic revision of partially preordered information and allows for implementing it with ASP. Finally, it shows how PPRSR can be applied to a real application of the VENUS european project.

Keywords: Partially preordered belief bases; Revision

Domains: Artificial Intelligence [cs.AI]

Mariette Serayet, Pierre Drap, Odile Papini. Extending Removed Sets Revision to partially preordered belief bases. International Journal of Approximate Reasoning, 2011, 52 (1), pp.110-126. ⟨10.1016/j.ijar.2010.05.008⟩. ⟨hal-01479553⟩