A Polygen Model for Heterogeneous Database Systems: The Source Tagging Perspective (VLDB ‘90)

Target Applications

• Heterogeneous database systems that present a unified view of data compiled from various local sources.

Goal:

• Identify “source tags” i.e. local sources (databases) from which data originated.
• Identify “intermediate tags” i.e. all sources (databases) which were used to derive a given output.

Contributions:

• Presents Polygen data model, built on top of the relational model.
• Describes the algebra underlying operations such as Select, Project, Join, Difference depicting how tags are propagated.
• Proposes a query translation mechanism for translating queries on Polygen tables to local tables.

Assumptions:

• Domain-mapping, attribute mapping, integration is done at priori. Considers only the problem of identifying tags.