Schema Evolution publication categorizer

Model management is a high-level programming language designed to efficiently manipulate schemas and mappings. It is comprised of robust operators that combined in short programs can solve complex metadata-oriented problems in a compact way. For instance, countless enterprise data integration scenarios can be easily expressed in this high-level language thus saving hundreds of development man-hours. Here we present the first model management engine that has reverse-engineering support for data integration, which is one of the most pressing metadata-oriented problems.

It is often difficult to obtain a good conceptual understanding of a legacy database, especially when there is a lack of documentation. Database reverse engineering attempts to provide solutions for this problem. It is the part of system maintenance work that produces a sufficient understanding of a legacy database and its application domain to allow appropriate changes to be made. However, research on database reverse engineering has largely ignored design and evaluation issues of their methods (i.e., foundations and processes).

The paper presents a DBMS-independent database reverse engineering
(DBRE) methodology based on a generic process model and on
transformation techniques. DBRE is proposed as a two-phase process
consisting in recovering the DBMSdependent data structures (data
structure extraction) then in recovering their semantics (data
structure conceptualization). The second phase, that is strongly
linked with the logical design phase of current database design
methodologies, can be performed by application of a selected set of
standard schema restructuring techniques, or schema transformations.

A methodology for extracting an extended Entity-Relationship (EER) model from a relational database is presented. Through a combination of data schema and data instance analysis, an EER model is derived which is semantically richer and more comprehensible for maintenance and design purposes than the original database. Classification schemes for relations and attributes necessary for the EER model extraction are derived and justified. These have been demonstrated to be implementable in a knowledge-based system; a working prototype system which does so is briefly discussed.