A New Taxonomy of Inconsistencies in UML Models : with their Detection Methods for Better MDE

MDE (Model Driven Engineering) is an emerging software engineering paradigm that relies on models as primary artifacts to build complex software and hardware systems. It basically aims at overcoming the inability of third-generation languages to alleviate the complexity of platforms and express domain concepts effectively. One particular vision of MDE was proposed by OMG (Object Management Group) under the name of MDA (Model Driven Architecture). MDA consists of a set of standardized rules and practices to promote good modeling and fully exploit the models in order to gain in sustainability, consideration of execution platforms and productivity. To achieve these objectives, the models to be used within an MDA approach should be consistent (i.e., complete, free of ambiguities, free of contradictions and anomalies, etc.). In this paper, we identify and describe possible inconsistencies that can be encountered in UML models while applying MDA, propose a new taxonomy for such inconsistencies and precise those supported by the most known verification approaches found in literature. Simple examples are used throughout the paper to illustrate our contribution.