REVER describes its products, which consist of DB-MAIN and various programming language analysers, as being about the development, maintenance, control and evolution of information systems, particularly with respect to the data contained within those information systems.
REVER is not just based on modeling techniques but employs a formal model-driven architecture. That is, it engineers (both forward and reverse) data in terms of a hierarchy of conceptual models, logical models and physical models. This is important because all databases share a common conceptual model and all databases of the same type share a common logical model.
The principles behind REVER are not new: reverse engineering and model-driven architectures have been around for a number of years. However, the two have not previously (as far as we know) been combined within a data-centric, as opposed to an application-centric, environment. The result is that REVER goes much further in automating the discovery of relevant structures than more conventional approaches.