利用XBRL技術設計可剖析的開放式企業報告架構

Using XBRL Technology to Design a Parsable “Open-ended” Business Report Framework

除了傳統財務報表資訊之外，企業財務報告中的附註資訊或是企業年報中包含的非財務性資訊、文字性資訊等「開放式資訊」，可能也「隱藏」了許多重要的企業資訊；然而，受限於資訊格式上的不一致，投資人必須透過繁瑣且錯誤發生率較高的人工處理程序，才能使用或流通此類資訊。因此，本研究認為應由資訊技術觀點，重新探討不同的企業報告資訊如何整合與塑模，以提升企業報告的整體品質。本研究嘗試應用設計科學研究方法，定義一個「開放式企業報告」資訊架構，再利用XBRL的「可尋獲的分類標準集合」技術將「開放式企業報告」中所包含的重要資料項目與文字加以標記化、模組化，並以公開發行公司各項企業報告資訊為案例，建置「開放式企業報告」所需之XBRL分類標準與分析技術，以驗證本研究所提出之概念性架構。另一方面，為了證實「XBRL格式化之開放性資訊」的相對優勢與用途，乃依據企業報告資訊品質之相關研究與理論，提出四個研究假說，並利用一個單因子受試者間的實驗設計，藉由研究發展出的「開放式企業報告」分類標準與案例文件等，於實驗室進行各項假說之測試；實驗結果顯示，對於一般非專業使用者而言，無論是資料的可靠性、可比較性、可取得性、攸關性與即時性等，可剖析、標準化與可延伸定義的XBRL「開放式企業報告」資訊，均較傳統非格式化、無系統性之企業報告具有較高的品質。

There exist a few problems with the current public-company’s informationdisclosure mechanisms. For example, due to the traditional close-designed framework,various types of business filings cannot be efficiently integrated to benefit informationusers. In addition, certain unparsable data format (e.g. PDF or HTML) leads to theunsearchable, unexchangeable and uninteroperable problems of data consumption.Therefore, using appropriate information technology (e.g. XBRL) to reconstruct theinformation disclosure mechanisms has drawn tremendous attention from marketparticipants for over a decade. This paper uses the design science methodology todevelop an “open-ended business report” framework. The framework adopts XBRLDTS technology that will allow owners of the business filings to integrate financial,non-financial reports more flexibly and provide higher search / analytical abilitywhile dealing with either formatted or unformatted data. On the other hand, to provethe proposed framework can increase the overall quality of business information,we conduct an experiment to test four hypotheses derived from related literatures.The experiment findings indicate the XBRL-based “open-ended” business reportframework can significantly improve reliability, comparability, accessibility,relevance and timeliness of business information. The empirical results suggest thecomparative advantages of proposed framework attribute to format parsability, formatstandardization and module extensibility of XBRL.