Information concerning the origin of data (that is, its provenance) is important in many areas, especially scientific recordkeeping. Currently, provenance information must be maintained explicitly, by added effort of the database maintainer. Since such maintenance is tedious and error-prone, it is desirable to provide support for provenance in the database system itself. In order to provide such support, however, it is important to provide a clear explanation of the behavior and meaning of existing database operations, both queries and updates, with respect to provenance. In this paper we take the view that a query or update implicitly defines a provenance mapping linking components of the output to the originating components in the input. Our key result is that the proposed semantics are expressively complete relative to natural classes of queries that explicitly manipulate provenance.
|Title of host publication||Database Theory – ICDT 2007|
|Subtitle of host publication||11th International Conference, Barcelona, Spain, January 10-12, 2007. Proceedings|
|Editors||Thomas Schwentick, Dan Suciu|
|Number of pages||15|
|Publication status||Published - 2007|
|Name||Lecture Notes in Computer Science|