Coupling solvers with model transformations to generate explorable model sets

Model transformation is an effective technique to produce target models from source models. Most transformation approaches focus on generating a single target model from a given source model. However, there are situations where a collection of possible target models is preferred over a single one. Such situations arise when some choices cannot be encoded in the transformation. Then, search techniques can be used to help find a target model having specific properties. In this paper, we present an approach that combines model transformation and constraint programming to generate explorable sets of models. We extend previous work by adding support for multiple solvers, as well as extending ATL, a declarative transformation language used to write such transformations. We evaluate our approach and language on a task scheduling case study including both scheduling constraints and schedule visualization.