On REA-based XBRL Versioning

With the coming of big data and the development of information technologies, it raiseshigher requests to the processing and reporting of corporation information. Now, more and morecompanies switch to XBRL (eXtensible Business Reporting Language). However, nowadays,with the rapid development of society, technologies, businesses, and accounting standardschange from day to day, so the same company will have different DTS (Discoverable TaxonomySet) and XBRL instance documents in different periods of time, possibly making datacomparability decrease and data analysis more difficult. How to build XBRL versioning in orderto validly finish the conversion and migration of the two different XBRL versions based on thesame entity is the focus of the current research of XBRL.In this paper, based on the analysis of REA model and XBRL Specification, the idea ofcombining REAand XBRL Versioning was proposed, and the XBRL Versioning Model was builtusing REA. The framework of XBRL versioning was constructed and the details including BaseModule, Concept Module (Concept Use Module, Concept Detail Module), Relationship SetModule and Dimension Module were interpreted by a three-tier hierarchy of Assignments,Actions, and Events. A detailed analysis of the processing of XBRL versioning including theinput, process, and the output of final versioning report was provided, distinguishing what can beautomated and what needs to be manual. Furthermore, the paper presented the minimalrequirements of conformance checking for DTSs and the versioning between different DTSs.On the one hand, the paper wants to provide a theoretical support for XBRL versioning andthe development of related software and tools; on the other hand it attemps to facilitate theintegration of REAand XBRL, exploring the future direction of accounting information system.