Information Extraction from Brazilian News Articles on Public Procurement Fraud

• Paulo Marcos Assis Universidade Federal de Santa Catarina (UFSC) https://orcid.org/0009-0004-9448-9982
• Jônata Tyska Carvalho Universidade Federal de Santa Catarina (UFSC)
• Márcio Castro Universidade Federal de Santa Catarina (UFSC)

DOI: https://doi.org/10.5753/sbbd.2026.249237

Resumo

Combating fraud in public procurement is a critical task for oversight agencies. While journalistic coverage can offers a rich source of early warning signals, its unstructured nature hinders automated auditing. To bridge this gap, we propose an Information Extraction (IE) pipeline to extract four key entities required for administrative linkage: municipality, procurement modality, notice number, and contract object. We evaluate three extraction approaches: (1) a Hybrid-Heuristic baseline; (2) a Retrieval-Augmented Generation (RAG) framework; and (3) an End-to-End Structured Large Language Model (LLM). Results demonstrate that the structured LLM significantly outperforms modular baselines, achieving an average accuracy of 91.45% across the four attributes. Furthermore, we introduce a gold-standard dataset of 796 annotated articles. All experiments are reproducible.

Palavras-chave: Information Extraction, Large Language Models, Retrieval-Augmented Generation, Public Procurement Fraud, News Articles

Referências

Assis, P. M. d., Castro, M., and Carvalho, J. T. (2025). Machine learning-based classification of portuguese news articles on public procurement fraud. In Anais da I Escola Regional de Aprendizado de Máquina e Inteligência Artificial da Região Sul (ERAMIA-RS 2025), Porto Alegre, RS. Sociedade Brasileira de Computação. Accepted; awaiting publication.

de Aquino, I. V., dos Santos, M. M., Dorneles, C. F., and Carvalho, J. T. (2024). Extracting information from brazilian legal documents with retrieval augmented generation. In Anais do IX Workshop on Data Science against Corruption in the Public Sector (DS-COPS 2024), pages 280–287, Florianópolis, SC, Brasil. SBC.

Hindi, M., Mohammed, L., Maaz, O., and Alwarafy, A. (2025). Enhancing the precision and interpretability of retrieval-augmented generation (rag) in legal technology: A survey. IEEE Access, 13:46171–46189.

Kariuki, P. and Reddy, P. (2017). Operationalising an effective monitoring and evaluation system for local government: Considerations for best practice. African Evaluation Journal, 5(2).

Lima, W., Lira, R., Paiva, A., Silva, J., and Silva, V. (2023). Methodology for automatic extraction of red flags in public procurement. In 2023 International Joint Conference on Neural Networks (IJCNN), pages 1–7, Gold Coast, Australia.

OECD (2018). The Role of the Media and Investigative Journalism in Combating Corruption. OECD Publishing, Paris.

Pereira, J., Assumpcao, A., Trecenti, J., Airosa, L., Lente, C., Cléto, J., Dobins, G., Nogueira, R., Mitchell, L., and Lotufo, R. (2025). Inacia: Integrating large language models in brazilian audit courts: Opportunities and challenges. Digital Government: Research and Practice, 6(1).

Premasiri, D., Ranasinghe, T., Mitkov, R., et al. (2025). Survey on legal information extraction: current status and open challenges. Knowledge and Information Systems, 67:1128–1135.

Schneider dos Santos, E., Machado dos Santos, M., Castro, M., et al. (2025). Detection of fraud in public procurement using data-driven methods: A systematic mapping study. EPJ Data Science, 14(52). Published 22 July 2025.

Silva, M., Santos, E., Alves, K., Silva, H., Pedrosa, F., Valencca, G., and Brito, K. (2024). Using generative ai for simplifying official documents in the public accounts domain. In Anais do XII Latin American Symposium on Digital Government (LASDIGOV 2024), pages 246–253, Brasília, DF, Brasil. SBC.

Sirqueira, E. J. V. and Vidal, F. d. B. (2024). Evaluation of named entity recognition using ensemble in transformers models for brazilian public texts. In Anais do XXI Encontro Nacional de Inteligência Artificial e Computacional (ENIAC 2024), pages 966–977, Belém, PA, Brasil. SBC.

Solanki, A., Main, H., Mehta, D., Kulkarni, D., and Dalvi, H. (2025). Lawbot: An enhanced legal information retrieval system using rag. In Anais da 2025 IEEE Silchar Subsection Conference (SILCON), pages 1–6, Dimapur, India. IEEE.

Souza, A. C. S. d. and Dorneles, C. F. (2025). Cono: Um coletor automatizado de notícias sobre corrupção em santa catarina. In Anais da XX Escola Regional de Banco de Dados (ERBD), pages 129–132, Florianópolis, SC. Sociedade Brasileira de Computação.

Souza, F., Nogueira, R., and Lotufo, R. (2020). Bertimbau: Pretrained bert models for brazilian portuguese. Intelligent Systems, pages 403–417.

UNODC (2014). Reporting on Corruption: A Resource Tool for Governments and Journalists. United Nations Office on Drugs and Crime, Vienna.

Wei, X., Cui, X., Cheng, N., Wang, X., Zhang, X., Huang, S., Xie, P., Xu, J., Chen, Y., Zhang, M., Jiang, Y., and Han, W. (2024). Chatie: Zero-shot information extraction via chatting with chatgpt.