Metamodels play a pivotal role in conceptual modeling as they manifest the abstraction level applied when creating conceptual models. Consequently, design decisions made by the metamodel developer determine utility, capabilities, and expressiveness of the conceptual modeling language - and eventually the created models. However, only limited research defines and applies metrics for analyzing the structure and capabilities of a metamodel, and eventually support the development of new metamodels. This not only concerns general-purpose modeling languages, but also domain-specific ones, which usually undergo shorter update cycles. The paper at hand introduces a generic analysis framework to syntactically analyze modeling languages. The framework is applied to 40 metamodels of domain-specific conceptual modeling languages (DSML). This research establishes a foundation to support metamodel development in the future. The contribution of this paper is threefold: i) an analysis framework for conceptual modeling method metamodels is proposed, ii) results from applying this framework to 40 ADOxx-based DSML metamodels are presented, and iii) a human-based reasoning after comparison of these results with Ecore-based metamodels is conducted.

Top