透過您的圖書館登入
IP:18.225.255.134
  • 學位論文

利用文字探勘技術建立XBRL財務報表電腦稽核系統之研究

Using The Text Mining Technology To Establish Computer Auditing System Of XBRL Financial Statements

指導教授 : 黃士銘
若您是本文的作者,可授權文章由華藝線上圖書館中協助推廣。

摘要


可延伸商業報導語言(eXtensible Business Reporting Language, XBRL)最重要的應用在於企業之財務報告書。我國行政院金融監督管理委員會在資本市場全球化的趨勢下,推動我國企業於2010年正式邁入XBRL財務報告申報元年。企業申報之XBRL四大財務報表及附註揭露皆以全球統一通行的標準格式為大眾所用,將使得更多財報使用者提高對企業資訊的關注,形成一個全民稽核的時代。此時,若企業無法掌握XBRL財報品質,除了讓舞弊行為有機可趁,更可能在風險發生時引發企業重大損失。 因此本研究旨在提出一個結合XBRL與資料探勘技術之財報檢核電腦稽核系統,除了能驗證財報上的數字資料,並運用資料探勘技術對文字性的附註揭露進行分析,藉以提高二者資料間判讀攸關能力與財報閱讀性,並透過這樣的功能設計讓使用者能夠易於進一步的查核與分析。依據本研究之實驗結果證明,本研究機制是可行、有效的並能提升稽核效率,冀望透過本研究機制能夠協助內部稽核人員提早或及時發現公司財報是否有重大錯誤與舞弊之可能,並能確實提高稽核效率與降低稽核時間與人力成本。 關鍵字:可延伸商業報導語言、電腦稽核、本體論、文字探勘、財務報表

並列摘要


The most important application of eXtensible Business Reporting Language (XBRL) is in financial statements of business. In the trend of globalization of capital markets, Taiwan’s Financial Supervisory Commission (FSC) promotes our businesses to use XBRL to report financial statements in 2010. The four financial statements and notes disclosure which enterprise reported are within a globally harmonized common standard format for use by the public, it will make the users of financial statements to improve more concern of business information; however, the era of everyone could be an auditor is coming. At this point, if the company is unable to get a good hold of the quality of XBRL finicail statements, it will not only increase the opportunities of fraud, but also may lead to significant losses while risks happened. Thus, the purpose of this study is to propose a combination of XBRL and data mining techniques to establish a computer auditing system to inspect XBRL financial statements, this mechanism not only able to verify numeric data but also uses text mining techniques to inspect notes disclosure which belong to text data type,which is in order to improve relevance of interpretation and statements readability between these two kinds of information, and let users could easier to further audit and analysis via the mechanism functions. Based on the experimental results of this study prove that the mechanism of this study is feasible, effective and can improve audit efficiency, hoping that through this mechanism could help internal auditors to early or timely detect if there is any possibility of serious errors and fraud, and help to improve the efficiency of audit and reduce audit time and labor costs. Keyword: XBRL, Computer Auditing, Ontology, Text Mining, Financial Statments

參考文獻


周濟群、連子杰,2011,運用文字探勘與XBRL技術提升企業資訊擷取與整合效益之研究,當代會計,第十二卷第一期:85-114。
陳稼興、謝佳倫、許芳誠,2000,以遺傳演算法為基礎的中文斷詞研究,資訊管理研究,第二卷 第二期(7月):27-44。
葉乃菁、王玳琪、張嘉珍、吳騏、賴志遠,2009,建構創新政策研究工具—文字探勘之應用簡介:科技政策研究與資訊中心。
劉艾華、王茂年、林英潔. (2009). 本體論及推論應用於資料語意偵錯之研究. 資訊管理學報.
林柄滄,2005,內部稽核理論與實務(第五版),台北。

延伸閱讀