ACQUA: Approximate consistent query answering over inconsistent knowledge bases

Consistent query answering is a principled approach for querying inconsistent knowledge bases. It relies on two central notions: a repair, that is, a maximal consistent subset of the facts in the knowledge base, and a consistent query answer, that is, a query answer entailed by every repair of the knowledge base. This paper presents the ACQUA system, which allows users to query inconsistent knowledge bases. Specifically, equality-generating dependencies are considered. Different from the standard notion of repair, where entire facts are deleted to restore consistency (which might lead to loss of useful information), the repair strategy adopted by ACQUA performs value updates within facts, thereby preserving more information in the knowledge base. An inconsistent knowledge base can admit multiple repairs; the ACQUA system computes a compact representation of all of them, called universal repair, which is also leveraged for query answering. Since consistent query answering is intractable in the considered setting, ACQUA implements a polynomial time algorithm to compute a sound (but not necessarily complete) set of consistent query answers.