Dealing with uncertainty in model transformations

Working with changing and vague/incomplete requirements is a challenging issue in Model Driven Engineering. Well known principles have been proposed to deal with the former case and reduce the effort required to implement a change. The second case requires managing uncertainty while expressing models and model transformations. Few research works have been proposed to express and manage models with uncertainty. However, uncertainty management in the model transformation itself remains a research gap. Hence, this paper introduces the notion of partial pattern which underpins the expression of uncertain model transformations. As part of the proposal, a transformation infrastructure that supports the execution of uncertain transformations is presented. Also, results of an evaluation of scalability are given. The performed experiments make use of application cases extracted from the Escape it! serious game. This is a learning game which intends to help autistic children reinforce and generalize visual skills.