A Domain Specific Modeling Language for REA

The Resource-Event-Agent (REA) ontology has its roots in the accounting discipline and was originally developed as a reference framework to conceptualize economic phenomena in an enterprise. In its proposal in 1982, McCarthy already had the vision to facilitate the design of data structures of accounting information systems by means of REA [1]. Since this time the REA model has been further extended and evolved into a domain ontology [2]. All REA concepts are based on well established concepts of the literature in economic theory which is certainly one of the strengths of REA. However, REA has no dedicated representation format and, consequently, no graphical syntax. Thus, users may struggle when describing the REA models leading to the impression that REA is a rather heavy-weight approach. McCarthy feels (as he expressed during our joint work in standardization activities) that a dedicated graphical syntax such as it exists for e3-value may help in overcoming this problem and may lead to a much more significant adoption of REA. Accordingly, we have started the endeavor of developing a domain specific modeling language for REA. 1 A domain-specific language (DSL) is a small, usually declarative language. It offers expressive power through appropriate notations and abstractions focused on – and usually restricted to – a particular problem domain [3]. Besides textual DSLs, we see an increasing interest in domain-specific modeling languages [4, 5]. According to [3] the benefits of a DSL approach are manifold: They allow solutions to be expressed at the level of abstraction of the problem domain. As a matter of fact, domain experts themselves can understand, validate and often modify DSL programs/models. The DSL programs/models are concise and self documenting to a large extent. They enhance productivity, reliability and maintainability. DSLs allow for validation and optimization at the domain level. When developing our REA-DSL we followed methodological steps that have been suggested by Strembeck and Zdun for the systematic development of domain specific languages [6]. Amongst other variants, they describe the development process for extracting a DSL from an existing system, which is appropriate