Accounting Fraud Detection Using Forensic Techniques Based on Sentiment Analysis and Interpretable Machine Learning : Focused on Internal Control over Financial Reporting

Woojune Jung; Jaewoon Yoon; Kyungho Kim

doi:10.24056/kar.2021.12.006

THE USE OF MACHINE LEARNING TECHNIQUES IN ACCOUNTING. A SHORT SURVEY

Journal of Social Sciences ◽

10.52326/jss.utm.2021.4(3).14 ◽

2021 ◽

Vol 4 (3) ◽

pp. 139-143

Author(s):

Mariana Vlad ◽

◽

Sorin Vlad ◽

Keyword(s):

Machine Learning ◽

Learning Algorithm ◽

Real Life ◽

Fraud Detection ◽

Machine Learning Algorithms ◽

Machine Learning Techniques ◽

Behavior Prediction ◽

Accounting Fraud ◽

Research Fields ◽

Learning Techniques

Machine learning (ML) is a subset of artificial Intelligence (AI) aiming to develop systems that can learn and continuously improve the abilities through generalization in an autonomous manner. ML is presently all around us, almost every facet of our digital and real life is embedding some ML related content. Customer recommendation systems, customer behavior prediction, fraud detection, speech recognition, image recognition, black & white movies colorization, accounting fraud detection are just some examples of the vast range of applications in which ML is involved. The techniques that this paper investigates are mainly focused on the use of neural networks in accounting and finance research fields. An artificial neural network is modelling the brain ability of learning intricate patterns from the information presented at its inputs using elementary interconnected units, named neurons, grouped in layers and trained by means of a learning algorithm. The performance of the network depends on many factors like the number of layers, the number of each neurons in each layer, the learning algorithm, activation functions, to name just a few of them. Machine learning algorithms have already started to replace humans in jobs that require document’s processing and decision making.

Download Full-text

INTERNAL CONTROL AND FRAUD PREVENTION: PRIOR RESEARCH ANALYSIS

Science and Studies of Accounting and Finance Problems and Perspectives ◽

10.15544/ssaf.2014.19 ◽

2014 ◽

Vol 9 (1) ◽

pp. 173-179

Author(s):

Stacey Mirinaviciene

Keyword(s):

Financial Reporting ◽

Internal Control ◽

Boards Of Directors ◽

Influencing Factor ◽

Fraud Detection ◽

Internal Controls ◽

Small Business Owners ◽

Prevention Measures ◽

Job Rotation ◽

Fraud Prevention

The focus of this study is to analyze prior research on fraud detection and prevention. Most researchers agree that strong internal controls are an influencing factor on fair financial reporting and fraud prevention and detection. Financial statement and employee fraud can be very expensive to businesses and the economy as a whole. The establishment and evaluation of the internal control methods and procedures can decrease fraudulent events and losses. Accounting professionals, CPA’s, and tax preparers are the first to detect “red flags” in business activities and must work together with boards of directors, CFO’s, and small business owners. Simple methods, such as ratio analyses can help to signal early signs of fraudulent events and prevent future damages. Implementation of fraud prevention measures are the most efficient deterrent. Some of the most effective controls like, job rotation, mandatory vacations, training, fraud hotlines, and surprise audits, need not be expensive and should be employed by all businesses. Unfortunately, the most important and effective fraud prevention techniques are seldom applied by businesses. Surprisingly, the least effective and most expensive measures, like external audits, are more frequently employed. As reported in this review of the literature, most businesses focus on fraud detection, while fraud prevention and implementing proper internal controls would result in better prevention of financial losses.

Download Full-text

PENGARUH EFEKTIVITAS PENGENDALIAN INTERN, ASIMETRI INFORMASI DAN KESESUAIAN KOMPENSASI TERHADAP KECENDERUNGAN KECURANGAN AKUNTANSI (Studi Empiris Kantor Pusat Bank Perkreditan Rakyat Di Kota dan Kabupaten Malang)

Adbis: Jurnal Administrasi dan Bisnis ◽

10.33795/j-adbis.v13i1.60 ◽

2019 ◽

Vol 13 (1) ◽

pp. 11

Author(s):

Restu Putri Pamungkas ◽

Istutik Istutik

Keyword(s):

Control System ◽

Information Asymmetry ◽

Financial Reporting ◽

Internal Control ◽

Profit Sharing ◽

Islamic Banks ◽

Accounting Fraud ◽

Internal Control System ◽

Information And Communication ◽

Study Offer

The research aims to examine the effect of the effectiveness of internal control on trends in accounting fraud, examine the effect of information asymmetry on trends in accounting fraud, examine the effect of compensation appropriateness on accounting fraud trends. Qualitative research was carried out through in-depth surveys and interviews with the management of Islamic banks in Malang, customers, and academics about the need for an internal control system that is in accordance with the characteristics of Islamic banks. The development of an internal control system using the COSO (Committee of Sponsoring Organizations of The Treadway Commissions) framework was adopted and adapted to the characteristics of Islamic banks. The results of the study offer a model of an Islamic bank's internal control system through five components of the COSO, namely (1) the control environment, (2) risk assessment, (3) control activities, (4) information and communication, and (5) monitoring that will benefit banks sharia to achieve strategic and operational goals, accuracy of financial reporting, and compliance with applicable policies and laws. Thus it will be able to help eliminate risks so that the performance of profit-sharing based financing can also be improved

Download Full-text

PREDICCIÓN DE FRAUDES EN EL CONSUMO DE AGUA POTABLE MEDIANTE EL USO DE MINERÍA DE DATOS

Universidad Ciencia y Tecnología ◽

10.47460/uct.v24i104.366 ◽

2020 ◽

Vol 24 (104) ◽

pp. 58-66

Author(s):

Fredy Humberto Troncoso Espinosa ◽

Fuentes Figueroa Paulina Gisselot ◽

Italo Ramiro Belmar Arriagada

Keyword(s):

Machine Learning ◽

Data Mining ◽

New York ◽

Electrical Power ◽

Fraud Detection ◽

Accounting Fraud ◽

Power Utility ◽

Palabras Clave ◽

Comprehensive Survey ◽

Prevention Methods

El comportamiento fraudulento en el consumo de agua potable es un problema importante que enfrentan las empresas de tratamiento de agua debido a que genera pérdidas económicas significativas. Caracterizar consumos fraudulentos es una tarea compleja, basada principalmente en la experiencia, y que presenta el desafío de la incorporación constante de nuevos clientes y la variación en el consumo mensual. En esta investigación, las técnicas de minería de datos se utilizan para caracterizar y predecir los consumos fraudulentos de agua potable. Para esto, se utilizó información histórica relacionada con el consumo. Las técnicas aplicadas mostraron un alto rendimiento predictivo y su aplicación permitirá enfocar eficientemente los recursos orientados a evitar este tipo de fraude. Palabras Clave: minería de datos, machine learning, agua potable, detección de fraude. Referencias [1]Centro de Investigación Periodística., «Producción y facturación de agua potable,» 30 Julio 2020. [En línea]. Disponible en: https://ciperchile.cl/wp-content/uploads/gestion-siis-2014-pag 88.pdf. [Último acceso: 30 Julio 2020]. [2]Bureau Veritas S.A., «https://www.bureauveritas.cl/es,» [En línea]. Disponible en: https://www.bureauveritas.cl/es/bureau-veritas-lider-mundial-en-ensayos-inspeccion-y-certificacion. [Último acceso: 1 Junio 2020]. [3]Essbio S.A., «www.essbio.cl,» [En línea]. [4]I. Monedero, F. Biscarri, J. Guerrero, M. Peña, M. Roldán y C. León, «Detection of water meter under-registration using statistical algorithms,» Journal of Water Resources Planning and Management, vol. 142, nº 1, p. 04015036, 2016. [5]I. Monedero, F. Biscarri, C. León, J. Guerrero, J. Biscarri y R. Millán, «Detection of frauds and other non-technical losses in a power utility using Pearson coefficient, Bayesian networks and decision trees,» International Journal of Electrical Power & Energy Systems, vol. 34, nº 1, pp. 90-98, 2012. [6]S. Wang, «A comprehensive survey of data mining-based accounting-fraud detection research,» de 2010 International Conference on Intelligent Computation Technology and Automation, New York, 2010. [7]J. Bierstaker, R. Brody y C. Pacini, «Accountants' perceptions regarding fraud detection and prevention methods,» Managerial Auditing Journal, vol. 21, nº 5, pp. 520-535, 2006. [8]C. Phua, V. Lee, K. Smith y R. Gayler, «A comprehensive survey of data mining-based fraud detection research,» arXiv preprint arXiv:1009.6119, 2010. [9]S. Kotsiantis, I. Zaharakis y P. Pintelas, «Machine learning: a review of classification and combining techniques,» Artificial Intelligence Review, vol. 26, nº 3, pp. 159-190, 2006. [10]J. Han, J. Pei y M. Kamber, Data Mining: Concepts and Techniques, Elsevier, 2011.

Download Full-text

Of Hurricanes and Harness Racing: The Accounting Fraud at DHB Industries, Inc.

Issues in Accounting Education ◽

10.2308/iace-50297 ◽

2012 ◽

Vol 28 (1) ◽

pp. 131-152

Author(s):

Michael C. Knapp ◽

Carol A. Knapp

Keyword(s):

Financial Reporting ◽

Internal Control ◽

Large Scale ◽

Audit Quality ◽

Securities And Exchange Commission ◽

Accounting Fraud ◽

Accounting Firms ◽

Fraud Risk ◽

Wide Range ◽

The Impact

ABSTRACT: This instructional case focuses on an accounting and financial reporting fraud involving DHB Industries, Inc., the nation's largest manufacturer of bullet-resistant vests. Three executives of this Securities and Exchange Commission (SEC) registrant, including its founder and CEO, masterminded a large-scale fraud that grossly misrepresented DHB's financial statements. The three executives colluded to conceal their misdeeds from the four accounting firms that served as the company's independent auditors over the course of the fraud. In late 2010, a federal jury convicted DHB's former CEO and COO of multiple counts of fraud and related charges. This case addresses a wide range of auditing issues raised by the DHB fraud, including the identification of fraud risk factors, auditing of related-party transactions, the impact of frequent auditor changes on audit quality, and the internal control reporting responsibilities of auditors.

Download Full-text

The Effects of Interaction between Internal Auditor and Audit Committee on Fraud Detection in Malaysia

International Journal of Engineering & Technology ◽

10.14419/ijet.v7i4.38.27820 ◽

2018 ◽

Vol 7 (4.38) ◽

pp. 1338

Author(s):

Sunita Lylia Hamdan ◽

Nahariah Jaffar ◽

Ruzanna Ab Razak

Keyword(s):

Financial Reporting ◽

Internal Control ◽

Audit Committee ◽

Specific Interaction ◽

Fraud Detection ◽

Internal Audit ◽

Internal Auditing ◽

Internal Auditor ◽

Internal Auditors ◽

External Auditors

This study aims to examine the effect of interaction between internal auditor and audit committee on fraud detection in Malaysia. Specific interaction is firstly; audit committee approving the appointment of chief audit executive, the evaluation of chief audit executive, the dismissal of chief audit executive, the internal audit budget and the internal audit plan or program. Secondly, audit committee’s involvement in reviewing internal auditor’s work specifically; providing input for the internal audit plan, reviewing the results of internal auditing related to financial reporting, reviewing the results of internal auditing related to internal control, reviewing the results of internal auditing related to compliance with laws and regulation, reviewing the internal audit involvement in management responses to internal audit suggestions, reviewing the difficulties or scope restrictions encountered by internal auditors and reviewing the coordination between internal auditors and external auditors. Survey questionnaires were mailed to internal auditors attached to 782 companies listed on Bursa Malaysia’s main market. The results of this study suggest that involvement of audit committee in approving chief audit executives’ matters is insignificant on internal auditors’ contribution to fraud detection. However, audit committee’s involvement in reviewing internal auditors’ work significantly influence the internal auditors’ contribution in fraud detection.

Download Full-text

Effects of CEO Turnover and Board Composition Reform on Improvements in the Internal Control Quality

International Journal of Financial Research ◽

10.5430/ijfr.v9n3p36 ◽

2018 ◽

Vol 9 (3) ◽

pp. 36

Author(s):

Hiroshi Uemura

Keyword(s):

Corporate Governance ◽

Financial Reporting ◽

Internal Control ◽

Chief Executive ◽

Audit Fees ◽

Internal Controls ◽

Board Independence ◽

Accounting Fraud ◽

Outside Directors ◽

The Government

Several serious accounting scandals have occurred in Japan in recent years (e.g., Olympus); however, the government, regulators, and auditing standard setters have struggled to identify new directions for corporate governance in listed companies, such as standard setting to address risks of fraud in an audit or the adoption of new corporate governance codes. The validity and effectiveness of monitoring by outside directors have received criticism within such a context. Nevertheless, in 2015, accounting fraud at Toshiba was discovered, which surprisingly involved upper management; the outside directors had failed to detect and prevent this fraud. Again, the monitoring function of the Japanese board of directors and outside directors was viewed with suspicion. Thus, this study examines Japanese corporations that disclose significant deficiencies (SDs) in internal controls over financial reporting (ICFR) and determines whether replacing the chief executive officer (CEO) and enhancing board members’ independence and financial expertise are followed by SD remediation. The results indicate that Japanese companies that disclose SDs in ICFR are more likely to replace their CEOs and enhance board independence. In addition, this study finds that although these actions do not affect SD remediation, upgrading the board’s accounting expertise does correlate positively with SD remediation. Moreover, if a company remediates a SD by increasing the number of accounting experts on the board, an increase in audit fees during the following term can be mitigated. These findings should be of interest to Japan’s regulators, auditing standard setters, and financial statement users when considering improvements in the quality of internal controls. In particular, these individuals must realize that the control environment is not improved in Japanese firms merely by replacing the CEO and increasing board independence, particularly because new CEOs encounter difficulties in changing the environment established by their predecessors.

Download Full-text

Fraud Specialists On Independent Audits

Journal of Business & Economics Research (JBER) ◽

10.19030/jber.v4i7.2686 ◽

2011 ◽

Vol 4 (7) ◽

Author(s):

James A. Tackett ◽

Fran M. Wolf ◽

Gregory A. Claypool

Keyword(s):

Financial Reporting ◽

Internal Control ◽

Fraud Detection ◽

Internal Controls ◽

Font Size ◽

Fraudulent Financial Reporting ◽

Background Checks ◽

Sarbanes Oxley ◽

Independent Auditors ◽

Accounting Software

The recent audit failures involving Enron, WorldCom, et al., have left the accounting profession and governmental regulators scrambling to find better methods of detecting and preventing fraudulent financial reporting. Congress passed the Sarbanes-Oxley Act of 2002 (SOX) which requires companies to report on the operating effectiveness of their internal controls over financial reporting.  Additionally, the independent auditor is required to assess and report on the effectiveness of their client’s internal controls, and they must attest to management’s internal control assessment.  Notably absent from SOX is a requirement that independent auditors must employ fraud specialists in their independent audits of SEC filers. This study examines the benefits and costs associated with requiring the use of fraud specialists on independent audits of SEC filers.  Fraud specialists have expertise better attuned to fraud detection not ordinarily possessed by regular auditors.  First, the narrow but deep perspective of the fraud specialist enables them to find fraudulent activity that would be missed by regular auditors.  Second, unlike regular auditors, fraud specialists employ methodologies that are effective in the presence of management collusion.  They are more highly skilled at interviewing potential witnesses and fraud suspects and are trained in recognizing deception.  Third, fraud specialists are better trained in the use of antifraud technology, methods, and computerized forensic accounting software.  Fourth, fraud specialists have superior investigative skills and can conduct covert examinations, access restricted databases, conduct background checks, and locate hidden assets better than regular auditors.  Finally, they understand the legalities of gathering evidence of fraud and can operate without violating the rights of potential witnesses and fraud suspects. Qualitative analysis demonstrates that utilizing fraud specialists on independent audits has positive net benefits to financial reporting.  Recommendations are made regarding the types of fraud detection/deterrence skills and techniques that would be beneficial to independent auditors.

Download Full-text

A Study on Various Applications of Data Mining and Supervised Learning Techniques in Business Fraud Detection

Machine Learning Applications for Accounting Disclosure and Fraud Detection - Advances in Finance, Accounting, and Economics ◽

10.4018/978-1-7998-4805-9.ch008 ◽

2021 ◽

pp. 108-125

Author(s):

Amit Majumder ◽

Ira Nath

Keyword(s):

Machine Learning ◽

Data Mining ◽

Supervised Learning ◽

Academic Research ◽

Fraud Detection ◽

Customer Relationships ◽

Financial Accounting ◽

Accounting Fraud ◽

Data Mining Technique ◽

Learning Techniques

Data mining technique helps us to extract useful data from a large dataset of any raw data. It is used to analyse and identify data patterns and to find anomalies and correlations within dataset to predict outcomes. Using a broad range of techniques, we can use this information to improve customer relationships and reduce risks. Data mining and supervised learning have applications in multiple fields of science and research. Machine learning looks at patterns of data and helps to predict future behaviour by learning from the patterns. Data mining is normally used as a source of information on which machine learning can be applied to solve some of problems in our daily life. Supervised learning is one type of machine learning method which uses labelled data consisting of input along with the label of inputs and generates one learned model (or classifier for classification type work) which can be used to label unknown data. Financial accounting fraud detection has become an emerging topic in the field of academic, research and industries.

Download Full-text

MORALITAS, PENGENDALIAN INTERNAL DAN GENDER DALAM KECENDERUNGAN TERJADINYA FRAUD

Jurnal Akuntansi ◽

10.24912/ja.v21i1.132 ◽

2017 ◽

Vol 21 (1) ◽

pp. 35

Author(s):

Luh Komang Merawati ◽

I Nym Kusuma Adnyana Mahaputra

Keyword(s):

Character Education ◽

Financial Reporting ◽

Internal Control ◽

Control Policy ◽

Accounting Fraud ◽

University Environment ◽

Economic Entity ◽

Funds Management ◽

And Gender ◽

The University

Morality, Internal Control and Gender in Fraud.University is an economic entity which is required to be accountable and transparent in funds management. Accountability can be indicated through financial reporting that is free from fraud. Fraud refers to the intentionally accounting errors to mislead the financial statement readers and usersin order to taking advantages. This research conducted with research design experiments 2x2 factorial by adding a gender perspective in examining the tendency of commit accounting fraud between individuals who have a (high or low) level of moral reasoning and (with or without) elements of internal control organization as factors that could cause accounting fraud. Data was analyze with two-ways ANOVA and Mann Whitney test in testing hypotheses. Research results indicate significant differences in the tendency of the accounting fraud due to the level of morality, internal control and gender. The results of this study are expected to provide recommendations in evaluating internal control policy to prevent fraudulent accounting practices and the importance of ethics and character education in the university environment to reduce the intention of fraud.

Download Full-text