（2007）Artificial Neural Network Assistant (ANNA) for Continuous Auditing and Monitoring of Financial Data.pdfVIP

JOURNAL OF EMERGING TECHNOLOGIES IN ACCOUNTING Vol. 4 2007 pp. 29–45 Artificial Neural Network Assistant (ANNA) for Continuous Auditing and Monitoring of Financial Data Eija Koskivaara Turku School of Economics and Turku Centre for Computer Science (TUCS) Barbro Back Abo Akademi University and Turku Centre for Computer Science (TUCS) ABSTRACT: This paper presents the development and applications of ANNA—Artificial Neural Network Assistant (ANNA) for continuous auditing and monitoring of financial data. The prototype presented here provides automatically six types of expectation values for account values and compares them to the actual values. In this study, ANNA is used for analyzing monthly account values. ANNA’s predictions or account expec- tations are based on function approximation (modeling) capability from the previous years’ monthly account values. ANNA uses a flexible one-step-ahead prediction model. The results are illustrated both by a graph on the computer screen and in the tables. The evaluation of results show that artificial neural networks technology is competitive compared to other methods used in this study. Keywords: analytical procedures; artificial neural network; continuous auditing. INTRODUCTION ecent audit failures and corporate scandals world-wide have brought to light the Rquestion of whether the auditing procedures are up-to-date and whether the use of analytical procedures (APs) or supervision of auditors are appropriate (Alles et al. 2004). During the past decades, the changes and challenges in the auditing environment have led to the publication