Abstract
Using a semantic data model, a program has been developed that takes as input a set of assertions about two database schemas and generates a schema for their
union. This schema is then used for either a virtual or physical representation of the original databases. In the case of a virtual merge, queries against the merged
databases are automatically decomposed. In the case of a physical merge, the data transfer programs are automatically generated. In either case, integrity constraints may be automatically generated to enforce consistency in places where the two databases contain overlapping information. The problem is the converse of providing '"user views": given two subschemas, what "super-schema" contains both?
union. This schema is then used for either a virtual or physical representation of the original databases. In the case of a virtual merge, queries against the merged
databases are automatically decomposed. In the case of a physical merge, the data transfer programs are automatically generated. In either case, integrity constraints may be automatically generated to enforce consistency in places where the two databases contain overlapping information. The problem is the converse of providing '"user views": given two subschemas, what "super-schema" contains both?
| Original language | English |
|---|---|
| Title of host publication | Proc IEEE Compcon Conference |
| Publisher | Institute of Electrical and Electronics Engineers |
| Pages | 279-286 |
| Number of pages | 8 |
| Publication status | Published - 1 Sept 1980 |