Deciding behaviour compatibility of complex correspondences between process models

Compatibility of two process models can be verified using common notions of behaviour inheritance. However, these notions postulate 1:1 correspondences between activities of both models. This assumption is violated once activities from one model are refined or collapsed in the other model or in case there are groups of corresponding activities. Therefore, our work lifts the work on behaviour inheritance to the level of complex 1:n and n:m correspondences. Our contribution is (1) the definition of notions of behaviour compatibility for models that have complex correspondences and (2) a structural characterisation of these notions for sound free-choice process models that allows for computationally efficient reasoning. We show the applicability of our technique, by applying it in a case study in which we determine the compatibility between a set of reference process models and models that implement them