Model-Driven Engineering and Elicitation Techniques: A Systematic Literature Review

Abstract

Model-Driven Engineering (MDE) aims to describe a system with the help of a series of models. More abstract models are refined into more concrete models by model transformations that ensure that the information from higher-level models is retained in lower-level models. This reduces the effort required to build lower-level models. The quality of lower-level models is also improved because they are partly based on well-tested model transformations. A key prerequisite for MDE is the definition of formal modeling languages, for which metamodeling is typically used. Formally defined modeling languages capture an ever-growing number of aspects of the software development process from requirements to architecture, design, and implementation details. A commonly held sentiment in the requirements modeling community is that requirements elicitation techniques have received less attention in terms of MDE. We perform a systematic literature review to either refute or confirm this sentiment by exploring the use of metamodels in papers on five well-known elicitation techniques: Cognitive Work Analysis, Complex Adaptive Systems Theory, Contextual Inquiry and Design, Distributed Cognition, and Emergent Knowledge Process design. The result of the survey indicates - within the limitations of the survey - that there is very little research applying metamodeling to the definition of these elicitation techniques.