The Nature of Data Reverse Engineering

Data Reverse Engineering is a specific domain of information system engineering aiming at rebuilding the documentation of legacy databases. The presentation explores this domain by presenting the basic problems, such as the implicit construct recovery, as well as techniques and tools for solving them. A reference process model is proposed, that identifies the main products, the classes of problems and their solving techniques. Two main processes are described, namely data structure extraction and data structure conceptualization. Through the first one, the complete logical schema (i.e., the programmer's view) is recovered, including both declared and implicit constructs and constraints. The data structure conceptualization process aims at interpreting the data structures in conceptual terms. Finally, the presentation describes a representative CASE environment that supports the processes by providing schema, data and code analysis tools as well as a rich schema transformation toolkit.