A NEW APPROACH TO LEVERAGING DOMAIN‐SPECIFIC MODELS

To appreciate how MIFs improve domain‐specific modeling tool‐chains, it is necessary to understand the processes implemented within model interpreters. Model interpreters implement model transformations. Model transformation is
one of the central activities in model‐driven development paradigms such as model‐driven architecture (MDA) and model driven engineering (MDE). Model transformation allows a single model to be used for a variety of purposes. Model
transformation also allows a “high‐level” language (like UML) to be automatically translated into a “low‐level” language (like C), just as compilers translate object‐oriented programs into machine code. In practical terms, a model interpreter is
usually implemented as a plug‐in to a modeling environment that performs operations on models created in that environment.

A model interpreter performs a semantic transformation on models. That is, a model interpreter performs a transformation that requires definition and understanding of the semantics, or meaning, of model elements. A model interpreter realizes the semantics of the modeling constructs on which it operates by defining the consequences of the use of those constructs within a given context. For example, translation between modeling languages whose constructs have possibly similar though not identical meanings is a semantic transformation. Translation of a graphical representation of a model into an equivalent textual form is not a semantic transformation.