Inconsistent knowledge-bases can entail useful conclusions when using the three-valued semantics of the paraconsistent logic LP. However, the set of conclusions entailed by a consistent knowledge-base under the three-valued semantics is smaller than set of conclusions entailed by the knowledge-base under a two-valued semantics. Preferring conflict-minimal interpretations of the logic LP; i.e., LPm, reduces the gap between these two sets of conclusions.

Preferring conflict-minimal interpretations introduces non-monotonicity. To handle the non-monotonicity, this paper proposes an assumption-based argumentation system. Assumptions needed to close branches of a semantic tableaux form the arguments. Stable extensions of the set of derived arguments correspond to conflict minimal interpretations and conclusions entailed by all conflict-minimal interpretations are supported by arguments in all stable extensions.