Towards Generic Modularization Transformations

Modularization concepts have been introduced in several modeling languages in order to tackle the problem that real-world models quickly become large monolithic artifacts. Having these concepts at hand allows for structuring models during modeling activities. However, legacy models often lack a proper structure, and thus, still remain monolithic artifacts. In order to tackle this problem, we present in this paper a modularization transformation which can be reused for several modeling languages by binding their concrete concepts to the generic ones offered by the modularization transformation. This binding is enough to reuse different modularization strategies provided by search-based model transformations. We demonstrate the applicability of the modularization approach for Ecore models.