XBRL Meta-Modeling Based On UML

XBRL(eXtensible Business Reporting Language) has been widely used in the world due to its prominent advantages on information submitting, analysizing and financial information communication. But XBRL specification is described in natural language, this might cause ambiguity for software developers. There is a need, therefore, to carry out graphical description of the technical specification to express its semantic more strictly. By this, XBRL International Standardization Committee(XSB) put forward the strategic action plan, and released the abstract model of 1.0 and 2.0. But the models are still lack of a complete expression of relational semantic. The definition of business rules(formula linkbase)are not clear, and there is no detail description of related resources and their relations.Based on meta-modeling framework, this thesis builds the meta-model to formalize the XBRL specifications using Unified Modeling Language(UML). Through the XBRL specification requirement analysis, this thesis puts forward three layer model of XBRL technology space based on the MOF and semantic triangle and defines the models in the technology space. On the basis of three-layer model, the semantic mechanism of XBRL specifications are also analyzed in this thesis from the perspective of object-oriented, the metamodel is built using StarUML and Data validation application of formula linkbase are carried out at the end of this thesis through assertion and formula.The main content and innovation include in:(1) add linkbase model which represents relation constraints to form a complete relationship expression, construct XBRL metamodel to better formalize the XBRL specification.(2) build an improved formula linkbase model by adding graphical description of the non-resources elements to formalize the XBRL semantic mechanism.(3)model the business rules through assertion and formula, verify the building and validating process of assertion and formula through the application of formula linkbase, and try to improve the XBRL abstract model and provide a foundation for developers to implement XBRL solutions.