A feature-based approach for variability exploration and resolution in model transformation migration

The key to success with Model-Driven Engineering is the ability to maintain metamodels and their related artifacts consistent over time. Metamodels can evolve under evolutionary pressure that arises when clients and users express the need for enhancements. However, metamodel changes come at the price of compromising metamodelrelated artifacts, including model transformations, necessitating their migration to again conform to the evolved metamodel. Restoring conformance of transformations is intrinsically difficult since a multitude of possible migration alternatives exist, which are unfeasible to be inspected manually. In this paper, we present an approach to explore variability in model transformation migration. Employing a feature-based representation of several possible transformation migrations, the approach permits modelers to explore and explicitly discover differences and conflicts among them. Once the desired migration alternatives are selected, the actual migration program is generated and executed by exploiting the EMFMigrate platform.