Skip to Main Content
Model driven architecture is an approach to develop software based on transforming models automatically. In this approach, models co-exist and can be modified independently in different stages of software development, but the consistency between models should be preserved by appropriate propagation of changes to the related models. However, most approaches do not support complete synchronization for uni-directional transformations. This paper proposes a new method for incremental synchronization of models that are related to each other by a uni-directional transformation. For this aim, traces of transformation are used for delete and update change propagation. Also, a knowledge base is generated automatically that encompasses patterns of transformation and employed for propagating insertion in source and target models. The knowledge base contains rules defined based on source and target metamodels that produced and evaluated automatically. The proposed method was evaluated by applying it to synchronize a class diagram as a source model and a relational database diagram as a target one.
Date of Conference: 13-14 Dec. 2011