Research On XBRL Financial Reporting And Accounting General Ledger: Taxonomy Improvement And Data Integration

XBRL (eXtensible Business Reporting Language) has been applied in more andmore countries and organizations in the world and has made a series of remarkableachievements, but a development bottleneck occurs in the current XBRL research: onthe one hand, XBRL Reporting Taxonomies still need further improvement; on theother hand, internal data also cannot "pass-through" XBRL financial reports. Standardimperfect and isolation of information inhibit enterprises’ enthusiasm and initiative toadopt XBRL technology, resulting in slow progress in application and promotion ofXBRL.The existing XBRL research lacks an overall perspective, and most of theresearch focuses on the study of XBRL financial reporting, leading to the limitation ofXBRL research. In this paper, based on the overall perspective of XBRL reportinginformation chain, to avoid over-concentration of XBRL research on a certain part ofinformation chain, in the meantime, the research is carried out around twofundamental issues:"how to improve the taxonomy of reports in each part ofinformation chain" and "how to integrate the data of each part of information chain".The settle of these problems would determine whether the application of XBRL canfit into enterprise application and move to a high level.This dissertation comprehensively employs the theories of economics,management， financial accounting, information science by involving theoretical andempirical methods and in the different stages of research using different researchmethods, and even a variety of methods combined. First, this dissertation reviews therelevant literature followed by the analysis of the current research and application ofXBRL, and further through theoretical analysis builds an information chain based on XBRL taxonomy, comparing different modeling methods in the information chainmodeling methods, and determining the corresponding method in different stage ofinformation chain. Due to the importance of technological factors, we give thedetailed analysis of the factors influencing the adoption of XBRL information chain.On this basis, this leads to two core issues-XBRL taxonomy improvement and dataintegration. In taxonomy improvement, the research is divided into XBRL generalledger taxonomy improvement and XBRL financial report taxonomy improvement,and the analyses are given respectively. In data integration, by building XBRL dataintegration framework, through the method of mapping and domain ontology, thedissertation studies the vertical data integration of XBRL accounting general ledgerand financial report, and the horizontal data integration of XBRL financial reports.Therefore the dissertation completes the research of XBRL taxonomy improvementand data integration.The main research conclusions are as follows:(1) The necessity of building of XBRL general ledger taxonomy module, so as toadapt to China’s national conditions. Based on such results, we build Chinese XBRLgeneral ledger taxonomy.XBRL general ledger taxonomy provides general standard of internal data, andevery country is expected to directly use the taxonomy. However, whether thetaxonomy is adapted to the Chinese context hasn’t been tested. By choosing China’snational accounting standards in the domain of accounting treatment as a benchmark,the dissertation evaluates the adaptability of international XBRL taxonomy. Althoughthe results show that the international taxonomy shows a strong level of completenessand both have adopted XML syntax specification, existing international taxonomy isnot directly applicable to China’s practical application. Through comparative analysis,the dissertation chooses to build Chinese XBRL general ledger taxonomy modulebased on international taxonomy and gives specific recommendations.(2) Through the analysis of the similarities in information disclosure among different industries, the efficiency and quality of the industrial taxonomy would beenhanced.In China, more and more industry extensions would be issued in future. However,due to the number of industry classification, the efficiency of building industryextension would become a problem. The dissertation excavates the characteristics ofthe industry information disclosure. Based on815XBRL instances of listed companyin Shanghai Stock Exchange in2011, through a matching between XBRL instancesand the corresponding SSE taxonomy, the findings of empirical analysis show thatthere are significant differences among different industries, but the differences do notexist among all industries. Then by cluster analysis method, according to thedisclosure situation of report items,21industry categories are grouped into7sets.Meanwhile, the findings show there are significant differences among groups for mostof the variables, and these differences are statistically significant. The method spansthe existing industry boundaries, thereby reducing the cost of industry extension andimproving the efficiency of the industry extension. In the meantime, the extension ofindustry taxonomies belonging to the same group can be carried out together, so thatit can improve the consistency and comparability of taxonomies, thereby enhancingthe quality of taxonomies.(3) The perspective of information chain is feasible for XBRL data integration.Current XBRL applications in the information chain are incomplete, and therewas a certain XBRL information island phenomenon. XBRL data integration includestwo aspects: on the one hand, the data integration between XBRL general ledger andfinancial report, on the other hand, the data integration between heterogeneous XBRLfinancial reports. Based on data integration theories, the paper attempts to buildXBRL data integration framework in the information chain, including longitudinaldata integration and horizontal integration of XBRL financial reports. Through theSRCD (Summary Reporting Contextual Data) module, achieving the link betweenaccounting general ledger and financial report, then it would make possible automatically generating XBRL financial reporting. And by building a unified domainontology of financial reporting, it provides a ontology basis for data integration, so asto fundamentally improve the efficiency of information integration. Through domesticXBRL application examples, the dissertation also validates the model of dataintegration of XBRL financial report. The results show that the scheme has certainfeasibility.Following are the primary innovation of this dissertation in four aspects:(1) Adopts and deepens the reporting information chain model based XBRLtaxonomy.The research from the perspective of XBRL information chain would help toavoid overly concentrated in a certain part of information chain and would take intoaccount more stakeholders’ interests. Most of previous XBRL research abouttechnology adoption is from the point of view of the regulator, rarely considering theadopter of XBRL technology. Through a detailed research of XBRL reportinginformation chain, the dissertation analyzes the micro-structure of information chainand the technological factors impacting the diffusion of XBRL reporting information chain.(2) Research on Chinese adaptability of international XBRL general ledgertaxonomy.Currently the research about the validity of international XBRL taxonomy inspecific countries is relatively rare. Based Chinese accounting treatment data interfacestandard, this is the first work to evaluate the adaptability of international XBRLgeneral ledger taxonomy, and to provide quantitative indicators. Based on theevaluation, the dissertation also provides a practical scheme for improvement.(3) Suggests the industry extension through cluster analysis of industrialinformation disclosureExisting literature show that there are significant differences among differentindustries, but there is no literature to further analyze whether this difference is causedby many industries or only a few industries. The findings show that the differences do not exist among all industries and the industry information disclosures show a certainsimilarity relations.Furthermore, through the cluster analysis, the industries showing similarcharacter of information disclosure are classified into the same group, thus to providea more effective extension scheme and enrich the theory of building taxonomy.(4) Suggests a unified ontology scheme of financial reporting domain.Most of the existing literature focuses only on ontology representation of a singleXBRL taxonomies. This approach can enhance the XBRL taxonomies semantics, butcan not fundamentally improve the efficiency of XBRL financial data integration.This paper presents unified financial reporting domain ontology, and builds theintegration model of XBRL financial reports. And taking as an example of domesticXBRL taxonomies, the conversion application is carried out.