Optimizing the Execution of Model Transformation Chain

Abstract

Model transformation is the key concept in the field of model-driven engineering. To scale up the model transformation tasks in a cloud-based low-code platform, reusing and composing smaller model transformations enables a complex transformation to be executed sequentially and optimally. To achieve reusing and composition of smaller transformations, we need to identify all the model transformation chains possible in the repository/folder to transform the source model to the target model. Then, with the help of static analysis, the total structural features involved in a model transformation are calculated and the chain with optimum quality criteria are considered to be the best chain. In this paper, we propose an approach that takes the benefit of static analysis and automatic program rewriting to optimize the selected chain. The approach will rewrite the transformations to keep only the rules and bindings that generate the intermediate model elements required to generate target model elements in the model transformation chain. The performance is improved by reducing the execution time and intermediate elements generated required to execute the selected transformation chain. This optimizing the chain' elements fasten the execution of the whole model transformation chain. The optimization of the execution of the model transformation chain helps in achieving the scalable solution that manages several workflows originating from heterogeneous model-driven platforms.