A model transformation classification method used in QVT

Aiming at potential pitfalls concerning interactivity and consistency of model driven development, this paper provides a new method of model transformation classification named MTCM. The method analyses the differences between the modeling language, the modeled system and the model itself, then makes a fine-granularity classification of model transformation based on the analyze. Afterwards, the classification is applied to the research of the model transformation standard Query/View/Transformation(QVT), which proposed by OMG together with Model Driven Architecture(MDA), and one of its classic example of the model transforming process from Ecore model to Kermeta model have been evaluated and verified critically. With the conclusion of analyzing QVT with the classification, we believe that understanding the MTCM method deeply is an important prerequisite for improving the veracity and efficiency of model transformation.