Title: Do data mining techniques assist auditors in predicting high-risk accounts in MENA Region countries?

Authors: Wafaa Salah Mohamed; Lamiaa Fattouh Ibrahim; Moid Uddin Ahmad

Addresses: Department of Business Administration, Faculty of Business Administration, Economics, and Political Science, The British University in Egypt, Cairo, Egypt ' Department of Computer Science, Faculty of Graduate Studies of Statistical Research, Cairo University, Cairo, Egypt; Faculty of Information Systems and Computer Science, October 6 University, Cairo, Egypt ' Lead Researcher and Faculty, Scholeio Education, NCR, India

Abstract: This study aims to construct a model that increases the accuracy of forecasting qualified audit opinions using publicly available measures and artificial intelligence. Additionally, the study probes the financial variables affecting an auditor's propensity to issue a qualified audit report. This study investigated the predictive abilities of three models: binary logistic regression, random forest, and decision tree. The study examined 564 audit reports (282 qualified reports) from nine MENA region countries from 2012 to 2018. The random forest technique produces the most accurate audit prediction. The study found that the significant firm-level variables that affect auditor opinion are book value per share, client size, and leverage ratio. The study's findings will bolster auditors, policymakers, and managers in effective decision-making.

Keywords: audit reports; decision tree; DT; random forest; logistic regression; qualified opinion.

DOI: 10.1504/AAJFA.2023.133421

Afro-Asian Journal of Finance and Accounting, 2023 Vol.13 No.5, pp.673 - 692

Received: 27 Jun 2021
Accepted: 07 Mar 2022

Published online: 15 Sep 2023 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article