A Rule-Based Language for Configurable N-way Model Matching

To build complex software-intensive systems, different stakeholders from diverse domains must collaborate to create and modify models. Model matching is a fundamental precondition of collaborative development, which is concerned with identifying common elements in input models. When stakeholders work on multiple models, they need to simultaneously compare all models to better understand differences and similarities. However, the literature shows no consensus on how to specify comparison criteria for matching multiple models, especially in a form that is independent of modeling language, which hampers their reuse and adoption. In this paper, we present a rule-based formalism that enables the user to specify their comparison criteria for multiple models at a high level of abstraction. We also introduce an N-way matching algorithm for comparing both homogeneous and heterogeneous models. As the tool support, we implemented a syntax-aware editor and a parser for specifying comparison rules for EMF-based models. The evaluation of our formalism shows that it is applicable in real modeling scenarios.