Résumé
Inconsistency handling is a fundamental problem in knowledge representation and reasoning. In this paper, we study this problem in the context of qualitative spatio-temporal reasoning, a framework for reasoning about space and time in a symbolic, human-like manner, by following an approach similar to that used for defining paraconsistent logics; paraconsistency allows deriving informative conclusions from inconsistent knowledge bases by mainly avoiding the principle of explosion. Inspired by paraconsistent logics, such as Priest’s logic LPm, we introduce the notion of paraconsistent scenario (i.e., a qualitative solution), which can be seen as a scenario that allows a conjunction of base relations between two variables, e.g., x precedes ∧ follows y. Further, we present several interesting theoretical properties that concern paraconsistent scenarios, including computational complexity results, and describe two distinct approaches for computing paraconsistent scenarios and solving other related problems. Moreover, we provide implementations of our two methods for computing paraconsistent scenarios and experimentally evaluate them using different strategies/metrics. Finally, we show that our paraconsistent scenario notion allows us to adapt to qualitative reasoning one of the well-known inconsistency measures employed in the propositional case, namely, contension measure.