Automated generation of structural design models from SBVR specification

Business rules play a critical role in developing effective and flexible information systems. In light of this critical role, the publication of the Semantic Business Vocabulary and Business Rules standard (SBVR) is considered a highly significant advance. However, although SBVR is a major breakthrough in defining business vocabulary and rules, its subsequent adoption in practical applications has proved to be much slower than originally anticipated. This paper proposes an approach that firstly reduces the SBVR complexity by identifying a subset that can be used to express structural design models unambiguously. Secondly, it formalizes a transformation that maps the identified SBVR subset to the UML class diagram, based on formally defined and validated meta-models that can guarantee the correctness of the generated UML class model. Finally, it incorporates the formal meta-models and the formal transformation algorithm into an automated transformation tool. The proposed approach is evaluated by taking into consideration the aspects of applicability, functionality, correctness, and completeness through a number of case studies and a comparison with an existing transformation approach.