ABox Abduction for Inconsistent Knowledge Bases under Repair Semantics

A new paper in computer science addresses ABox abduction for inconsistent knowledge bases (KBs) under repair semantics. ABox abduction seeks extensions of a KB to entail a non-entailed fact, with applications in diagnosis, explainability, and repair. While well-studied for consistent KBs, little is known for inconsistent ones, often caused by erroneous data. The authors define abduction notions and criteria for useful hypotheses, using repair semantics to regain meaningful reasoning. They provide a complexity landscape for ABox abduction under repair semantics for the lightweight description logics DL-Lite and EL_bot.