Title: A systematic literature review of machine learning techniques in financial fraud prevention and detection

Authors: Selorm Kofi Tagbo; Adebayo Felix Adekoya

Addresses: Department of Computer Science and Informatics, University of Energy and Natural Resources (UENR), P.O. Box 214, Sunyani, Bono Region, Ghana ' Department of Computer Science and Informatics, University of Energy and Natural Resources (UENR), P.O. Box 214, Sunyani, Bono Region, Ghana

Abstract: This review is aimed at analysing published articles in the field of fraud prevention and detection to determine the quantum of classification works that have been done using either supervised or unsupervised machine learning techniques. Again, by analysing these papers methodically, we would be able to identify the year that records the highest fraud publications and further rate continents based on the number of publications. Six hundred and sixty-two papers published in online databases like Scopus, Google Scholar, Web of Science, and Microsoft Academic between 2010 and 2022 were initially retrieved. Snowballing helped to discover more relevant articles. The search was inspired by PRISMA. Results showed that supervised machine learning technique was widely used as compared to the unsupervised learning counterpart. The number of publications in this field increased greatly between 2020 and 2021. Also, it was revealed that financial fraud cases within the African continent received the least attention from researchers.

Keywords: machine learning; financial fraud; financial fraud detection; artificial intelligence.

DOI: 10.1504/IJSSS.2023.135915

International Journal of Society Systems Science, 2023 Vol.14 No.4, pp.303 - 348

Received: 19 Dec 2022
Accepted: 03 May 2023
Published online: 09 Jan 2024 *

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