Model composition in Model Driven Engineering: A systematic literature review

Context: Model Driven Engineering (MDE) aims to alleviate complexity and improve reusability in software development. The development of complex software implies to divide it into independent parts before then assembled. This is how the problem of model composition has become an interesting and stills an emerging topic in MDE. Objective: Our goal is to analyze the current state of the art in model composition in the context of Model Driven Engineering. Method: We use the systematic literature review based on the guidelines proposed by Biolchini et al., Brereton et al., and Kitchenham and Charters. We propose five research questions and six quality assessments. Results: Of the 9270 search results, 56 have been considered relevant studies. These studies have resulted in 36 primary studies. Conclusion: The evaluation shows that most of approaches allow more than two models as inputs of the composition, allow composing heterogeneous models and enable the tuning of the composition schema, while the important limitations are about the maturity of implementations and the lack on the management of future evolutions or backwards compatibility.