Provenance Analysis for Missing Answers and Integrity Repairs

Data provenance approaches track how the answer to a database query derive from input items; however, prior approaches used “positive” provenance and were not directly usable for explaining “expected” but missing answers. A similar problem arises with the failure of integrity constraints. Our perspective is to offer explanations via possible (minimal) repairs using provenance. This is useful for debugging, repairing, and cleaning databases. In this paper, we introduce a novel approach to this problem for both missing/erroneous answers and integrity failures. The approach uses recent advances in provenance for first-order model checking.