A Relational Approach to Model Transformation with QVT Relations Supporting Model Synchronization

Kun Ma (University of Jinan, China)

Bo Yang (University of Jinan, China)

Zhenxiang Chen (University of Jinan, China)

Ajith Abraham (Scientific Network for Innovation and Research Excellence, USA)

Abstract: With the help of model transformation, it is possible to generate target models from source models. A possible way to face iterative development process with frequent modifications is to use not only a single transformation but also frequent model synchronization. In this paper, we propose a relational approach to model transformation using Query/View/Transformations (QVT) Relations language that also provides model synchronization mechanism based on the version of the models. The proposed framework uses a Platform-Independent Business Model (PIM-BM) and a Platform-Specific Business Component Model (PSM-BC) via the extension of the UML metamodel and MOF at different levels of abstraction, which sufficiently describe both the structural and behavioral properties of generic Web applications. Also we present the typical model mapping rules between PIM-BMs and PSM-BCs using QVT Relations. Finally the model synchronization based on the version of models is provided for the above model transformation approach.

Keywords: model driven software development, model synchronization, model transformation, model-driven architecture, modeling

Categories: D.2.1, D.2.11, D.2.2, I.6.5