Start Date

16-8-2018 12:00 AM

Description

Conceptual modeling languages such as BPMN and UML are widely used in industry and academia. Such modeling languages are usually introduced in overarching specification documents maintained by standardization institutions. Being the primary - often even the single - source of information, such specifications are vital for modelers, researchers, and tool vendors. However, how to derive a coherent and comprehensive specification was never systematically analyzed. This paper reports on the analysis of 11 current modeling language specifications with a focus on how metamodels are specified. Identified metamodel specification techniques are discussed and their sample usage is illustrated. Thereby, individual strengths and weaknesses of each technique are discussed. The contribution of this paper is a foundation for increasing the consistency and expressive power of metamodel specifications, ultimately leading to an improved understanding and better utilization of modeling languages.

Share

COinS
 
Aug 16th, 12:00 AM

How are Metamodels Specified in Practice? Empirical Insights and Recommendations

Conceptual modeling languages such as BPMN and UML are widely used in industry and academia. Such modeling languages are usually introduced in overarching specification documents maintained by standardization institutions. Being the primary - often even the single - source of information, such specifications are vital for modelers, researchers, and tool vendors. However, how to derive a coherent and comprehensive specification was never systematically analyzed. This paper reports on the analysis of 11 current modeling language specifications with a focus on how metamodels are specified. Identified metamodel specification techniques are discussed and their sample usage is illustrated. Thereby, individual strengths and weaknesses of each technique are discussed. The contribution of this paper is a foundation for increasing the consistency and expressive power of metamodel specifications, ultimately leading to an improved understanding and better utilization of modeling languages.