The Xbrl Technology Based On Semantic Web Model And Its Application Research

XBRL (eXtensible Business Reporting Language) is a markup language which is used to define standards of releasing financial and operational reports based on some operating standard. Nowadays, XBRL has been widely supported and applied in more than20countries and regions including America, Japan and China etc. During the application of XBRL,XBRL International gradually developed a series of technical specifications to meet the new demands of applications, and there are17technical specifications overall such as XBRL Base Specification, Dimensions Specification, Formula Specification, Versioning Specification and Inline XBRL (Rendering)Specification, etc. However, these technical specifications show some disadvantages, such as lack of inadequate capacity to reveal semantic meanings, safety issue, reasoning disability and lack of theoretical base. Meanwhile, Semantic Web, which appeared just the same time as XBRL, has the advantages of explicit semantics, reasoning mechanics, security design and solid theoretical basis. Therefore, it is an important research topic to improve XBRL a series of specification s which is a support for the whole application of XBRL with the semantic web technologyThis dissertation makes a systematic and synthetically research and analysis on the domestic and international status quo, the necessity and feasibility, specific improvement method and effectiveness of improving XBRL specifications with Semantic Web technology.Firstly, the XBRL Technical Model on Semantic Web basis (XTMSW) is proposed.According to the disadvantages from in-depth analysis of a series of existing XBRL technical specifications and the advantages of Semantic Web, this dissertation introduces the Semantic Web to the XBRL specifications. The XBRL Base Specification, Dimensions Specification, Formula Specification, Versioning Specification and Inline XBRL(Rendering)Specification are studied with Semantic Web technology. Then the XTMSW is proposed and deeply analyzed from7levels and6model views. The levels are Coding Level, Syntax Level, Resource Description Level, Ontology Level, Logic Level, Security Level, User Level, and the views are Conceptual Model View, Logical Model View, Development Model View, Operating Model View, Data Model View and Physical Model View.Secondly, XTMSW-based XBRL taxonomies are constructed.The comparisons among ontology construction methods are made. On the basis of XTMSW and Chinese Accounting Standard for Business Enterprises, Combined with the comparison results and the current XBRL taxonomies developing methods, this dissertation proposes the method and steps of constructing the XTMSW-based taxonomies. The methods and steps are exemplified by the taxonomy construction of Chinese Accounting Standard for Business Enterprises No.1-Inventories. Subsequently, this dissertation compares and analyzes the XTMSW-based inventories taxonomies and the existing inventories taxonomies from the construction range, the theoretical basis and the aspect of programming. Finally, relevant recommendations about constructing XTMSW-based taxonomies of Chinese Accounting Standard for Business Enterprises are proposed.Thirdly, the XTMSW-based XBRL Instance Document Retrieval System Prototype is developed.After studying the developing process of retrieval model, this dissertation selects the Semantic Retrieval Model for the retrieval of XTMSW-based Instance Documents. Combined with the generally used framework of retrieval system, the retrieval model framework of XTMSW-based Instance Documents is built. Then, after an elaboration of the semantic inference core algorithms, the prototype of XTMSW-based Instance Document Retrieval System is developed, and also the comparison between the instance documents searching results generated from the existing inventories taxonomies and XTMSW-based inventories taxonomies is made. Finally, the XTMSW-based instance documents retrieval system is evaluated by the indicators such as Recall and Precision.This paper draws the following main conclusions:First, it is becoming the trend to reform XBRL technical framework with the semantic web technology; Second, it has strong feasibility of building the XBRL technical model based on Semantic Web(XTMSW); Third, the scope of Purpose Taxonomy of Chinese Accounting Standards for Business Enterprises should be extended to the whole Chinese Accounting Standards for Business Enterprises including of recognition, measurement and disclosure; Fourth, the instance retrieval system based on XTMSW is better than the current instance retrieval system.