The Chase&Backchase algorithm for rewriting queries using views is based on constructing a canonical rewriting candidate called a universal plan (during the chase phase), then chasing its exponentially many subqueries in search for minimal rewritings (during the backchase phase). We show that the backchase phase can be sped up significantly if we instrument the standard chase to maintain provenance information. The particular provenance flavor required is known as minimal why-provenance in the literature, and it can be computed by exploiting the analogy between a chase step execution and query evaluation. © 2013 Springer-Verlag Berlin Heidelberg.
CITATION STYLE
Deutsch, A., & Hull, R. (2013). Provenance-directed chase&backchase. Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 8000, 227–236. https://doi.org/10.1007/978-3-642-41660-6_11
Mendeley helps you to discover research relevant for your work.