DefinePK

Abstract

This study explores how advanced analytics enhances fraud detection in forensic accounting, addressing the growing complexity of financial fraud. Employing a secondary qualitative approach, the research synthesises literature and analyses three case studies to evaluate analytics techniques, their practical outcomes, and implementation challenges. Findings reveal that machine learning achieves 85% accuracy in detecting anomalous transactions, while data mining and natural language processing uncover fraud patterns and deceptive communications effectively. Results from case studies show major impacts, e.g., $8 million restitution in a corporate fraud case, 40% loss reduction in a cyberfraud case, and policy reforms in a public sector scam. Yet, adoption is hindered by challenges including poor data, skill shortage and ethical concerns. It highlights that analytics can be an agile and accurate approach to fraud detection in forensic accounting, a process that otherwise takes time. It suggests hybrid models, where human expertise is mixed and recommends educating and setting ethical guidelines for responsible scaling of analytics. This research helps provide practitioners, policy makers, and academics with the ability to minimise fraud using analytics.