Interactive visual analysis of anomalous accounts payable transactions in SAP enterprise systems

Abstract

Purpose – The purpose of this paper is to demonstrate the technical feasibility of implementing multi-view visualization methods to assist auditors in reviewing the integrity of high-volume accounting transactions. Modern enterprise resource planning (ERP) systems record several thousands of transactions daily. This makes it difficult to find a few instances of anomalous activities among legitimate transactions. Although continuous auditing and continuous monitoring systems perform substantial analytics, they often produce lengthy reports that require painstaking post-analysis. Approaches that reduce the burden of excessive information are more likely to contribute to the overall effectiveness of the audit process. The authors address this issue by designing and testing the use of visualization methods to present information graphically, to assist auditors in detecting anomalous and potentially fraudulent accounts payable transactions. The strength of the authors ' approach is its capacity for discovery and recognition of new and unexpected insights. Design/methodology/approach – Data were obtained from the SAP enterprise (ERP) system of a real-world organization. A framework for performing visual analytics was developed and applied to the data to determine its usefulness and effectiveness in identifying anomalous activities. Findings – The paper provides valuable insights into understanding the use of different types of visualizations to effectively identify anomalous activities. Research limitations/implications – Because this study emphasizes asset misappropriation, generalizing these findings to other categories of fraud, such as accounts receivable, must be made with caution. Practical implications – This paper provides a framework for developing an automated visualization solution which may have implications in practice. Originality/value – This paper demonstrates the need to understand the effectiveness of visualizations in detecting accounting fraud. This is directly applicable to organizations investigating methods of improving fraud detection in their ERP systems.

Big Data Opportunities for Accounting and Finance Practice and Research

“…() use visual analytics to monitor financial stability. In the management accounting field, visual analytics identifies anomalous and potentially fraudulent accounts payable transactions (Singh and Best ).…”

Data‐driven auditing: A predictive modeling approach to fraud detection and classification

“…Audits are highly complex social processes, and subject to the conflict of divergence of interests between shareholders and owners, which might affect auditor performance and independence (Arya & Glover, ; Suddaby, Gendrom, & Lam, ). Singh and Best () demonstrated the technical feasibility of implementing multiview visualization methods to assist auditors in reviewing the integrity of high‐volume accounting transactions. In connection to these findings and our prior discussion of the principal‐agent dilemma and the underlying risk of moral hazard, we argue that an approach that helps to analyze, identify, and visualize data in a better way reduces information asymmetry by making the data more accessible to both the principal and agent.…”

Data visualization and cognitive biases in audits

“…It has been demonstrated that the human brain can process more information at a faster rate when the information is presented graphically than when it is presented in text or tabular formats (Lurie and Mason, 2007). Graphic displays allow auditors, especially novice auditors, to quickly comprehend the data, detect certain relevant patterns and identify trends and relationships that would take much more effort to identify if presented in tables (Vessey and Galletta, 1991;Singh and Best, 2016). Thus, visualization can be used in every phase of an audit, including decisions to accept or continue an audit engagement, audit planning, risk assessment, responding to risk assessments and tests of controls or substantive testing (AICPA, 2015).…”

“…Currently, visualization is being used by some auditors to generate insights, increase the accuracy of conclusions and improve the efficiency and effectiveness of the auditing process (Singh et al, 2013;AICPA, 2015;Brown-Liburd et al, 2015;Cao et al, 2015;Singh and Best, 2016) [1]. Data visualization can be used to both explore and explain data (Alawadhi, 2015).…”

“…Prior research (Joyce and Biddle, 1981;Lurie and Mason, 2007) shows that these biases can be operationalized through the manipulation of reference points, the format of the graphic presentation, or the vividness and evaluability of the data. Based on the insights from audit practitioners and regulators as well as from extant psychology and behavior research (Benbasat and Dexter, 2986;Hammersley, 2006;Lurie and Mason, 2007;KPMG, 2011;Singh and Best, 2016;Rose et al, 2017), six approaches to debiasing auditors' visualization analyses are identified: using data visualization to complement rather than supplement traditional audit evidence; positioning data visualization to support rather than replace sophisticated analytics tools; using a dashboard with multiple dimensions; using both visualized and tabular data in analyses; assigning experienced audit staff; and providing pre-audit tutorials on cognitive bias and visualization.…”