The Rule-based Classification for the Embezzlement Detection using Case-based Reasoning

Abstract

Internal embezzlement is a major source of corporate losses. Companies are making various efforts to detect, prevent, and mitigate it. Embezzlement detection involves identifying financial irregularities through red flag behaviors, such as secretive work habits and unused vacation time. It also involves reviewing employee spending using machine learning-based transaction data analytics and forensic analysis. Furthermore, robust controls, such as double authorizations and regular audits, help detect issues such as fraudulent vendors, fake invoices, and misuse of company cards, thereby preventing significant financial losses. In this paper, we discuss a unique hybrid classification method for detecting embezzlement of corporations. We expect the proposed hybrid classification method to increase the accuracy of corporate embezzlement detection.