INTEGRATING REENGINEERED DATABASES TO SUPPORT DATA FUSION

Large information systems often require the fusion of multiple databas es to achieve desired functionality. In this article, we focus on how automated query formulation capabilities may be realized over a set of fused databases. Reengineering issues related to database design and fusion are discussed, and a query formulation and design system known as QUICK is described. A case study is presented in which the logical schemas for two independent U.S. Army databases are reverse engineered into conceptual schemas that are subsequently used for data fusion an d automatic query generation. In addition, enhanced methods that emplo y meta-level conceptual constructs to support reverse engineering, dat a fusion, and query formulation are described.