Detecting Fraud in Public Procurement: A GMM-Based Approach to Analyzing Tender Data

• Fernando Augusto Schmitz Universidade Federal de Santa Catarina (UFSC) / Ministério Público de Santa Catarina (MPSC)
• Lívia Ferrão Universidade Federal de Santa Catarina (UFSC)
• Matheus Machado dos Santos Universidade Federal de Santa Catarina (UFSC)
• Márcio Castro Universidade Federal de Santa Catarina (UFSC)
• Jônata Tyska Carvalho Universidade Federal de Santa Catarina (UFSC)

DOI: https://doi.org/10.5753/sbbd.2024.240649

Resumo

Corruption and bid rigging in public procurement distort competition and increase the costs of products and services for public institutions, causing problems in different societal domains. The current availability of public data in digital format brings opportunities for applying machine learning to build solutions that help to deal with corruption. However, there are many challenges, like data sparsity and modeling complexity. Furthermore, confirmed cases of fraudulent tenders are limited, making applying traditional supervised learning techniques unfeasible. This work proposes a novel methodology for analyzing patterns using Gaussian Mixture Models (GMM) to identify suspicious bidding patterns when only a few fraudulent cases are known. Our methodology tests the similarity of unlabeled tenders, which can be fraudulent or not, with the fraudulent cases in different subspaces for defining a risk indicator. We run experiments in a dataset with tender data for acquiring heavy equipment purchases in which only a few cases are known as fraudulent. Results showed that our GMM-based methodology effectively provides a risk indicator ranking, highlighting risky tenders, making it a valuable tool for public agencies to enhance transparency and accountability in procurement.

Palavras-chave: Public procurement fraud, Gaussian Mixture Models (GMM), Machine learning in procurement, Anomaly detection in tenders, Government transparency initiatives

Referências

Goryunova, N., Baklanov, A., and Ianovski, E. (2021). Detecting corruption in singlebidder auctions via positive-unlabelled learning. HSE University.

International, T. (2022). Corruption perception index. Acesso em: 26 abril 2023.

Ministério Público de Santa Catarina (2022a). Mpsc desarticula esquema de propina para compra de máquinas pesadas em santa catarina. [link]. Accessed: 30 Sept 2023.

Ministério Público de Santa Catarina (2022b). Operação patrola: ex-prefeito de tangará é condenado a 13 anos de prisão. [link]. Accessed: 30 Sept 2023.

Rabuzin, K. and Modrušan, N. (2019). Prediction of public procurement corruption indices using machine learning methods. In Proceedings of the 11th International Joint Conference on Knowledge Discovery, Knowledge Engineering and Knowledge Management (IC3K 2019), pages 333–340. SCITEPRESS.

Rothstein, B. (2011). The Quality of Government: Corruption, Social Trust, and Inequality in International Perspective. University of Chicago Press.

Tribunal de Contas do Estado de Santa Catarina (2024). e-sfinge: Sistema de fiscalização integrada de gestão. [link]. Accessed: 26 Apr 2024.

Velasco, R. B., Carpanese, I., Interian, R., Paulo Neto, O. C. G., and Ribeiro, C. C. (2021). A decision support system for fraud detection in public procurement. International Transactions in Operational Research, 00:1–21.