Specific Areas Of The Formalization Of The Modeling Language And Its Model Consistency Validation Studies

Domian-Specific Modeling(DSM) is a Model-Driven Software Development methodology for the specific domain, regarding specific domain analysis and modeling as its core and the domain model as the main products of software development. Therefore, in the Model-Driven Software Development for specific domain, quality of software product is decided by quality of domain model. Domain model's consistency as one of the basic and most important properties of domain model quality and its validation, which are much less mature and deep than UML in the research and application, have become research hotspot in the field of DSM.Due to its the informal definition, Domian-Specific Meta-Modeling Language and modeling Language can not strictly describe its structure characteristics and constraint relationships and can not validate the consistency of its own too, so model consistency based on them can only be validated by hard-coding or empirical rules, resulting in the lack of holistic, systematic and accuracy of validation.In response, based on our preliminary study for Domian-Specific Meta-Modeling Language-XMML(XML-Based Meta-Modeling Language) and its support environment, according to expansion and refinement of XMML for domain model's consistency and its validation, this dissertation presents the formal definitions based on first-order logic of the abstract syntax and structural semantics of XMML, based on this, and establishes a formal framework for metamodel based on first-order logic in the domain and a definition mechanism of domain consistency constraints based on metamodel's symbol set, and then proposes concepts and validation methods based on first-order logical inference of consistency of metamodel in the meta-domain and consistency of metamodel in the domain and consistency and validity of domain consistency constraints and consistency of model in the domain, and then studies automatic mapping mechanism of formalisation of metamodel and model to a group of first-order logic formulas and designs and implements the corresponding automatic mapping prototype tool, and finally does case studies with DSM in the software architecture domain and computer network topology domain as examples respectively to show feasibility and practicality of the approach. As the study of the applied theory for validation of metamodel and model based on formalisation of meta-modeling language in the field of DSM, this dissertation greatly promotes further researching of domain model transformation and domain codes making. So, it has some theoretical significance and practical value.