On provenance minimization

Abstract

Provenance information has been proved to be very effective in capturing the computational process performed by queries, and has been used extensively as the input to many advanced data management tools (e.g. view maintenance, trust assessment, or query answering in probabilistic databases). We study here the core of provenance information, namely the part of provenance that appears in the computation of every query equivalent to the given one. This provenance core is informative as it describes the part of the computational process that is inherent to the query. It is also useful as a compact input to the above mentioned data management tools. We study algorithms that, given a query, compute an equivalent query that realizes the core provenance for all tuples in its result. We study these algorithms for queries of varying expressive power. Finally, we observe that, in general, one would not want to require database systems to evaluate a specific query that realizes the core provenance, but instead to be able to find, possibly off-line, the core provenance of a given tuple in the output (computed by an arbitrary equivalent query), without rewriting the query. We provide algorithms for such direct computation of the core provenance. ABSTRACTProvenance information has been proved to be very effective in capturing the computational process performed by queries, and has been used extensively as the input to many advanced data management tools (e.g. view maintenance, trust assessment, or query answering in probabilistic databases). We study here the core of provenance information, namely the part of provenance that appears in the computation of every query equivalent to the given one. This provenance core is informative as it describes the part of the computational process that is inherent to the query. It is also useful as a compact input to the above mentioned data management tools. We study algorithms that, given a query, compute an equivalent query that realizes the core provenance for all tuples in its result. We study these algorithms for queries of varying expressive power. Finally, we observe that, in general, one would not want to require database systems to evaluate a specific query that realizes the core provenance, but instead to be able to find, possibly off-line, the core provenance of a given tuple in the output (computed by an arbitrary equivalent query), without rewriting the query. We provide algorithms for such direct computation of the core provenance.

Ariadne

“…For example, consider the interval encoding (Figure 9(b)) for the example network shown in Figure 9(a). The provenance of tuple 3:1 is represented as a single interval [1,4], because the TID-Set forms a single contiguous sequence of TIDs (1 to 4). Interval encoding is most advantageous for queries involving aggregations over long sequences of contiguous TIDs, but introduces overhead if such sequences do not occur -both the start and end TID of an interval need to be stored.…”

“…Consider the covering intervals shown in Figure 9(b). The TID-Set for tuple 3:1 is covered by the interval [1,4]. This covering interval is stored in two additional fields at the end of the data tuple 3:1 .…”

Semiring-annotated data

“…2 We call such data K-annotated unordered XML, or simply K-UXML. Given a domain L of labels, the usual mutually recursive definition of XML data naturally generalizes to K-UXML: 3…”

“…In particular, Amsterdamer et al [2] studied provenance minimization, and defined the core of provenance information, namely the part of provenance that appears independently of the query plan that is in use. 6 They also provided algorithms that, given a query, compute an equivalent one that realizes the core provenance for all tuples in its result, as well as algorithms to compute the core provenance of a particular tuple in a query result without re-evaluating the query.…”

“…Since the publication of these two papers on semiringannotated relations and XML, a number of followup studies have investigated semantics and provenance models for additional query operators [14,3,4,17] and data models [31], the interaction of provenance information with query optimization [15,26], minimization [2] and factorization [29] of provenance expressions, and other topics. Moreover, K-relations have been incorporated Table hop represents possible links between nodes in a network.…”