本文關鍵詞： 財務共享服務模式 智能審計 大數據 深度學習　出處：《重慶理工大學》2017年碩士論文　論文類型：學位論文

【摘要】：大型集團對集中管理和數據共享的強烈需求,使得財務共享服務中心在跨國公司和國內大型集團企業(yè)中掀起了應用的熱潮。財務共享服務模式和傳統(tǒng)財務管理模式之間有較大區(qū)別,管理模式的改變給內部審計帶來了全方位的影響。面對財務共享服務模式,內部審計部門應該如何有條不紊地開展內部審計工作?如何更高效率且高質量完成內部審計數據分析工作?這些都是亟需解決的理論和現實問題。本文運用會計學、審計學、財務管理、計算機應用技術等學科理論,以信息論、系統(tǒng)論作為理論基礎和思維向導,充分發(fā)揮財務共享服務中心的優(yōu)勢,結合大數據的特點和人工智能技術,構建了財務共享服務模式下企業(yè)內部智能審計路徑。并開發(fā)了內部審計數據分析系統(tǒng)。首先,本文從四個步驟,即建設智能審計信息系統(tǒng)、內部審計人員的組織管理、智能審計實施程序、獲取審計證據,一步一步構建出財務共享服務模式下企業(yè)內部智能審計路徑。在智能審計實施程序中,提出通過智能審計分析進行審計預警,發(fā)現審計疑點;并圍繞審計疑點采用MECE分析法確定審計重點;再通過審計知識庫提供的審計方案,利用非現場審計為主和現場審計為輔的取證模式及時地獲取充分、可靠、相關的審計證據。其次,由于利用人工智能技術是實現智能審計的重點,本文把智能審計路徑中利用智能數據分析發(fā)現審計疑點作為單獨一章進行重點討論。在此部分,在智能審計分析中采用對比分析法,通過從企業(yè)內外部的大數據中挖掘出可靠且具有參考價值的比較基準值,與被審單位值對比分析,進行審計預警,從而發(fā)現審計疑點。同時,由于財務共享服務模式下的審計數據具備大數據“4V”((Volume、Variety、Velocity、Veracity)特點,在審計數據集市里使用人工操作的方式調用數據進行分析,會造成調用數據不精準且審計效率低的局面。因此,本文通過深度學習網絡構建了智能審計數據分析模型,采用自動編碼器的原理使數據自學習,主動識別和提取非結構化數據的數字和文本特征,再通過訓練數據與選取審計項目下的評價指標進行一一對應,使得審計數據智能匹配到審計項目下進行數據的對比分析,提高審計效率和質量。最后,基于理論部分的分析,利用MATLAB 2015a開發(fā)平臺、SQL Anywhere9.0數據庫,設計并開發(fā)了內部審計數據分析系統(tǒng),同時結合案例的形式印證了智能審計數據分析模型的可行性。
[Abstract]:The strong need for centralized management and data sharing in large groups, The financial sharing service center has set off an upsurge of application in multinational corporations and domestic large group enterprises. There is a great difference between the financial sharing service mode and the traditional financial management mode. The change of the management mode has brought the omnidirectional influence to the internal audit. Facing the financial sharing service mode, how should the internal audit department carry out the internal audit work methodically? How to complete internal audit data analysis efficiently and qualitatively? These are theoretical and practical problems that need to be solved. This paper uses the theories of accounting, auditing, financial management, computer application technology and so on, and takes information theory and system theory as the theoretical basis and thinking guide. This paper gives full play to the advantages of financial sharing service center, combines big data's characteristics and artificial intelligence technology, constructs the internal intelligent audit path of enterprise under the mode of financial sharing service, and develops the internal audit data analysis system. This article from four steps, namely, constructs the intelligent audit information system, the internal auditor organization management, the intelligent audit implementation procedure, obtains the audit evidence, In the implementation program of intelligent audit, it is put forward that the audit warning is carried out through intelligent audit analysis, and the audit doubt is found. MECE analysis method is used to determine the audit focus around the audit doubt points, and then through the audit plan provided by the audit knowledge base, the evidence collection mode of off-site audit and on-the-spot audit is used to obtain sufficient and reliable evidence in time. Secondly, because the use of artificial intelligence technology is the focus of intelligent audit, this paper focuses on the use of intelligent data analysis to find audit suspects in the intelligent audit path as a separate chapter. In the intelligent audit analysis, by using the comparative analysis method, through excavating the reliable and reference reference value of the comparison benchmark value from the inside and outside of the enterprise big data, comparing and analyzing with the value of the audited unit, carrying on the audit early warning, At the same time, because the audit data under the financial sharing service mode has the characteristics of big data "4V" Volumey Varietyn Velocity verity, the audit data is transferred in the audit data set market by manual operation for analysis. Therefore, this paper constructs the intelligent audit data analysis model through the deep learning network, and adopts the principle of automatic encoder to make the data self-learning. The digital and text features of the unstructured data are identified and extracted actively, and then the training data and the evaluation index under the audit item are matched one by one, so that the audit data can be intelligently matched to carry out the comparative analysis of the data under the audit project. Finally, based on the theoretical analysis, the internal audit data analysis system is designed and developed based on the MATLAB 2015a development platform. At the same time, the feasibility of intelligent audit data analysis model is confirmed by case study.
【學位授予單位】：重慶理工大學
【學位級別】：碩士
【學位授予年份】：2017
【分類號】：F626.115;F239.4

【參考文獻】

相關期刊論文 前10條

1 馮玲;;試探財務共享服務模式下的大數據審計[J];會計之友;2017年06期

2 陽杰;應里孟;;大數據時代的審計證據與審計取證研究[J];財會月刊;2017年01期

3 劉星;牛艷芳;唐志豪;;關于推進大數據審計工作的幾點思考[J];審計研究;2016年05期

4 殷瑞剛;魏帥;李晗;于洪;;深度學習中的無監(jiān)督學習方法綜述[J];計算機系統(tǒng)應用;2016年08期

5 程平;崔納牟倩;;大數據時代基于財務共享服務模式的內部審計[J];會計之友;2016年16期

6 鄭偉;張立民;楊莉;;試析大數據環(huán)境下的數據式審計模式[J];審計研究;2016年04期

7 程平;白沂;;基于財務共享服務模式的大數據審計研究[J];中國注冊會計師;2016年05期

8 李強;謝汶莉;;大數據審計中的可視分析[J];中國內部審計;2016年02期

9 陳偉;SMIELIAUSKAS Wally;;大數據環(huán)境下的電子數據審計:機遇、挑戰(zhàn)與方法[J];計算機科學;2016年01期

10 孫s，

本文編號：1505393