The Representation Problem in Reverse Engineering

Building models to understand software systems is an important part of reverse engineering. Formal and explicit model building is important because it focuses attention on modeling as an aid to understanding and results in artifacts that may be useful to others. The representation used to build models has great innuence over the success and value of the result. Choosing the proper representation during reverse engineering is the representation problem. This paper examines the representation problem by presenting a taxonomy of solutions. It also illustrates the issues involved in choosing a representation through an example reverse engineering task.