Storage and Use of Provenance Information for Relational Database Queries

Abstract

In database querying, provenance information can help users understand where data comes from and how it is derived. Storing the provenance data is critical in the sense that, the storage cost should be as small as possible and of fine granularity, and it should support the user query on provenance tracking efficiently as well. In this demo, we have implemented a relational database system prototype which can support SQL-like query while supporting provenance data recording during query execution. In particular, we propose a tree structure to store provenance information and further propose various reduction strategies to optimize its storage cost; we support the functionality of provenance data tracking at tuple level for user queries in a visualized way.