Bibliometric analysis of big data applications in accounting fraud detection

Abstract

Purpose: The use of Big Data can be strategic in the prevention and detection of accounting fraud, helping to protect the integrity of accounting and improving the efficiency of processes. Using these advanced technologies, organisations can minimize their exposures to fraud and maintain more accurate and transparent accounting. Design/methodology/approach: A bibliometric study was conducted, using a quantitative method based on indicators and on an analysis of the links and connections and visualisation of the network to see the impact, citation, institutions, authors, countries, etc., as well as collaboration between the various agents. Findings: Research on Big Data in fraud detection has grown rapidly, integrating Machine Learning, Data Mining, and Artificial Intelligence into auditing. The field is global and interdisciplinary, shifting from traditional anomaly checks to predictive, real-time prevention while acknowledging emerging digital fraud risks. Research limitations/implications: The selection from the Scopus database was made using 6 keywords; the results could have been different if a different or larger set of keywords had been applied to the research carried out. Practical implications: The findings provide valuable insights into the academic landscape surrounding this topic. Advanced techniques such as Data Mining, Machine Learning, and Deep Learning can significantly enhance transaction transparency and traceability, thereby bolstering confidence in accounting practices and mitigating fraud risks. Social implications: Fraud is inherently unethical and presents significant challenges for finance and society; technology plays an important role in its detection and prevention. However, reinforcing ethical and sustainability principles is increasingly important. This issue transcends mere technological application; it requires fostering a culture of integrity, transparency, and responsibility. Originality/value: Through an integrated bibliometric approach, this study provides original value by mapping how Big Data is reshaping fraud-detection research and by highlighting the persistent gap between advanced analytics and audit-standard alignment, underscoring the need for explainable and audit-ready tools.