PROV-IDEA: Supporting Interoperable Schema and Data Provenance within Database Evolution

Database evolution and data provenance are two closely related research fields. On the one hand, the registry (via provenance) of the schema evolution allows the maintenance of its version record. On the other hand, the origin of the data (i.e., its provenance) will always be affected by modifications (i.e., the evolution) in the schema on which they are based. Despite these interrelationships, there are few works in the literature that have proposed advances in that direction. In particular, to the best of our knowledge, there is no research that has resulted in a general and interoperable solution to the problem of managing database evolution using provenance. In this article, we present PROV-IDEA: a PROV-Interoperable Database Evolution Approach. This is a proposal that allows the simultaneous management of the provenance of schemas (of relational databases) and data, using the PROV standard as a way to guarantee interoperability. Furthermore, it is an adaptable and expandable approach (by using PROV templates), which allows a non-intrusive and seamless integration with existing applications, as well as different aspects of provenance information generation. These properties are demonstrated in the article by presenting a proof of concept built on top of a third-party relational database evolution tool.