面向XBRL的财务报表聚类分析审计系统研究-软件工程专业毕业论文.docxVIP

I I 摘 要 具有正确性、完整性、合法性和公允性的企业财务报表数据，对一国社会经 济生活产生重要影响，加强对企业财务报表审计是现代经济健康、有序发展的必 然要求。然而，面对电子化水平极高的企业财务会计账务处理系统和电子化数据， 审计人员过去采用的传统手工条件下的查账方法和审计模式，因受到审计手段、 审计时间和成本的制约，无法对海量的财务数据进行全面审查，使当前审计工作 的质量和效率受到严峻挑战。 面对上述情况，本课题依据《审计法》和其它审计相关法规，结合财务报表 审计的基本原理和具体审计方法程序，针对当前较流行的 XBRL 财务报表审计系统 开发需求，展开面向 XBRL 的财务报表聚类分析审计系统的研究。在对企业和单 位财务报表的构成项目、要素进行标准化研究的基础上，利用会计记账原理形成 的会计报表各项目数据之间的内在依存和对应关系，通过数据聚类算法，查出不 符合依存和对应关系的项目数据，及时发现财务报表的错报行为，提高财务信息 的正确性。 本课题在进行大量需求分析的前提下，针对当前 XBRL 财务报表编制模式， 运用聚类算法对数据进行挖掘、分析的功能，使用 Borland Delphi7.0 和 Microsoft SQL Server 2000 开发工具，设计了一个应用于当前审计工作的通用审计系统。通 过本文提出的技术体系和实现方案，开发的审计系统可以应用于审计工作实践， 帮助审计人员分析企业海量的电子财务数据，查找和发现异常数据，形成审计证 据和审计工作底稿，从而制止企业财务报表造假及违法乱纪现象，提高审计效率， 促进审计事业和国民经济的可持续、健康发展。 关键词：XBRL，数据挖掘，聚类，审计 II II ABSTRACT Accuracy, completeness, legality, authenticity of corporate financial statements data has an important influence on the economic and social life of a country. It is a necessity to strengthen the auditing of corporate financial statements for a healthy and orderly development of modern economy. However, faced with the ever-increasing higher level computerization of management system and electronic data of corporate financial accounting, auditors find it difficult to have a complete examination of a good amount of financial data with the traditional method of accounting examination and auditing modes which is restricted by the means, time and cost of auditing. This poses a real challenge to the quality and efficiency of the work of auditing. To solve these problems, this thesis first makes a systematic study on the clustering analysis auditing system of XBRL-oriented financial statements according to the development demand for the current XBRL financial statements. Then, by analyzing the constitutive items, elements of corporate and unit financial statements on the basis of a standardized study, this thesis makes an attempt to find out the item data which are not in conformity with the said dependency and correspondence by means of digging the various i