Reverse Engineering of Relational Database Schema based on Universal Metadata Queries

The paper presents an approach to reverse engineering of a relational database schema. The approach enables database schema extraction from the corresponding metadata stored in the data dictionary by executing a universal set of queries regardless of the source database management system. The concrete queries are generated by combining predefined templates with concrete metadata for the particular source system. Based on the proposed approach, we implemented a software tool and experimentally validated it. The implemented solution enables reverse engineering of relational databases managed by several contemporary database management systems.