THE INFLUENCE OF DATA ENGINEERING ON DATA QUALITY FOR MACHINE LEARNING MODEL TRAINING: A NARRATIVE LITERATURE REVIEW

Authors

  • Carlos Eugênio Mendes Paes Landim
  • Mauro José Araújo de Melo

DOI:

https://doi.org/10.63330/sasciencesv6n2-323

Keywords:

Algorithmic bias, Data Drift, Data Leakage, Data quality, Machine Learning

Abstract

This article analyzes the extent to which data quality techniques applied to the development of Machine Learning models contribute to mitigating biases and risks associated with automated decision-making. A narrative literature review was conducted, based on the purposive selection of fourteen studies according to explicit inclusion and exclusion criteria and an extraction protocol with direct verification of the original sources. The results indicate that data quality techniques have a direct application in addressing statistical and operational problems, allowing issues related to data preparation, model evaluation validity, and changes in data distributions observed in production to be identified and reduced. The evidence analyzed also indicates that these techniques are not sufficient, when applied in isolation, to mitigate social biases and risks associated with automated decisions: cases involving automated recruitment and criminal risk assessment show that apparently neutral attributes can reproduce relationships associated with sensitive characteristics, while spurious correlations may cause models to rely on patterns that do not adequately represent the phenomenon of interest. As a practical contribution, a data quality checklist is proposed, organized according to the stages of the Machine Learning lifecycle, in which each check is traceable to the section and source that support it. The study concludes that data quality should be understood more broadly, incorporating, in addition to statistical criteria, aspects related to representativeness, data curation, production context, auditing, and continuous monitoring.

Downloads

Download data is not yet available.

References

ANGWIN, Julia et al. Machine bias. ProPublica, New York, 23 maio 2016. Disponível em: https://www.propublica.org/article/machine-bias-risk-assessments-in-criminal-sentencing. Acesso em: 31 out. 2026.

BAROCAS, Solon; SELBST, Andrew D. Big data's disparate impact. California Law Review, v. 104, n. 3, p. 671-732, 2016. Disponível em: https://escholarship.org/content/qt31b8c97t/qt31b8c97t.pdf. Acesso em: 31 out. 2026.

BARROS, Kawan Ribeiro Dias; MELO, Mauro José Araújo de. Impacto da qualidade de dados públicos tabulares no desempenho de modelos clássicos de Machine Learning. RevistaFT, v. 29, ed. 152, nov. 2025. DOI: https://doi.org/10.69849/revistaft/ra10202511300622. Disponível em: https://revistaft.com.br/impacto-da-qualidade-de-dados-publicos-tabulares-no-desempenho-de-modelos-classicos-de-machine-learning/. Acesso em: 1 set. 2026.

BATISTA, Gustavo Enrique de Almeida Prado Alves. Pré-processamento de dados em aprendizado de máquina supervisionado. 2003. 232 f. Tese (Doutorado em Ciências de Computação e Matemática Computacional) – Instituto de Ciências Matemáticas e de Computação, Universidade de São Paulo, São Carlos, 2003. DOI: https://doi.org/10.11606/T.55.2003.tde-06102003-160219. Disponível em: https://teses.usp.br/teses/disponiveis/55/55134/tde-06102003-160219/pt-br.html. Acesso em: 1 set. 2026.

BATISTA, Gustavo Enrique de Almeida Prado Alves; PRATI, Ronaldo C.; MONARD, Maria Carolina. A study of the behavior of several methods for balancing machine learning training data. ACM SIGKDD Explorations Newsletter, v. 6, n. 1, p. 20-29, 2004. Disponível em: https://doi.org/10.1145/1007730.1007735. Acesso em: 7 nov. 2026.

BORGES, Gustavo Silveira; FALEIROS JÚNIOR, José Luiz de Moura. Viés racial em sistemas de inteligência artificial: desafios para a responsabilidade civil e a necessidade de parametrização do risco. Revista IBERC, Belo Horizonte, v. 6, n. 2, p. 100-128, 2023. DOI: https://doi.org/10.37963/iberc.v6i2.237. Disponível em: https://revista.iberc.org.br/iberc/article/view/237. Acesso em: 1 set. 2026.

BRECK, Eric et al. The ML test score: a rubric for ML production readiness and technical debt reduction. In: IEEE INTERNATIONAL CONFERENCE ON BIG DATA, 2017, Boston. Proceedings [...]. Piscataway: IEEE, 2017. p. 1123-1132. DOI: https://doi.org/10.1109/BigData.2017.8258038. Disponível em: https://research.google/pubs/pub46555/. Acesso em: 31 out. 2026.

CHAWLA, Nitesh V. et al. SMOTE: synthetic minority over-sampling technique. Journal of Artificial Intelligence Research, v. 16, p. 321-357, 2002. Disponível em: https://doi.org/10.1613/jair.953. Acesso em: 25 set. 2026.

DASTIN, Jeffrey. Amazon scraps secret AI recruiting tool that showed bias against women. Reuters, San Francisco, 10 out. 2018. Disponível em: https://www.reuters.com/article/us-amazon-com-jobs-automation-insight-idUSKCN1MK08G. Acesso em: 31 out. 2026.

DWORK, Cynthia et al. Fairness through awareness. In: INNOVATIONS IN THEORETICAL COMPUTER SCIENCE CONFERENCE, 3., 2012, Cambridge. Proceedings [...]. New York: ACM, 2012. p. 214-226. DOI: https://doi.org/10.1145/2090236.2090255. Disponível em: https://arxiv.org/abs/1104.3913. Acesso em: 31 out. 2026.

GAMA, João et al. A survey on concept drift adaptation. ACM Computing Surveys, v. 46, n. 4, art. 44, p. 1-37, 2014. Disponível em: https://doi.org/10.1145/2523813. Acesso em: 31 out. 2026.

GEBRU, Timnit et al. Datasheets for datasets. Communications of the ACM, v. 64, n. 12, p. 86-92, 2021. Disponível em: https://doi.org/10.1145/3458723. Acesso em: 31 out. 2026.

GRANT, Maria J.; BOOTH, Andrew. A typology of reviews: an analysis of 14 review types and associated methodologies. Health Information & Libraries Journal, v. 26, n. 2, p. 91-108, 2009. Disponível em: https://doi.org/10.1111/j.1471-1842.2009.00848.x. Acesso em: 31 out. 2026.

JAKUBIK, Johannes et al. Data-centric artificial intelligence. Business & Information Systems Engineering, v. 66, n. 4, p. 507-515, 2024. DOI: https://doi.org/10.1007/s12599-024-00857-8. Disponível em: https://link.springer.com/article/10.1007/s12599-024-00857-8. Acesso em: 1 set. 2026.

JESUS, Ananda Fernanda de; BARITÉ, Mario; SANTAREM SEGUNDO, José Eduardo. O termo qualidade de dados na Ciência da Informação brasileira: uma análise de terminologia pontual. In: ENCONTRO NACIONAL DE PESQUISA EM CIÊNCIA DA INFORMAÇÃO, 24., 2024, Vitória. Anais [...]. Vitória: ANCIB, 2024. p. 1-16. Disponível em: https://www.enancib.ancib.org/enancib/en/article/view/615. Acesso em: 1 set. 2026.

KAPOOR, Sayash; NARAYANAN, Arvind. Leakage and the reproducibility crisis in machine-learning-based science. Patterns, v. 4, n. 9, 100804, 2023. DOI: https://doi.org/10.1016/j.patter.2023.100804. Disponível em: https://www.sciencedirect.com/science/article/pii/S2666389923001599. Acesso em: 1 set. 2026.

LARSON, Jeff et al. How we analyzed the COMPAS recidivism algorithm. ProPublica, New York, 23 maio 2016. Disponível em: https://www.propublica.org/article/how-we-analyzed-the-compas-recidivism-algorithm. Acesso em: 31 out. 2026.

MITCHELL, Margaret et al. Model cards for model reporting. In: CONFERENCE ON FAIRNESS, ACCOUNTABILITY, AND TRANSPARENCY, 2019, Atlanta. Proceedings [...]. New York: ACM, 2019. p. 220-229. Disponível em: https://doi.org/10.1145/3287560.3287596. Acesso em: 31 out. 2026.

ROCHA, Lucas Helfstein. Mitigating drift in machine learning systems through continuous input monitoring: an architectural proposal and empirical evaluation of detection methods. 2026. 98 p. Dissertação (Mestrado em Ciência da Computação) – Instituto de Matemática, Estatística e Ciência da Computação, Universidade de São Paulo, São Paulo, 2026. DOI: https://doi.org/10.11606/D.45.2026.tde-12052026-175759. Disponível em: https://teses.usp.br/teses/disponiveis/45/45134/tde-12052026-175759/en.html. Acesso em: 1 set. 2026.

ROSA, Alexandre Morais da; GUASQUE, Bárbara. Inteligência artificial, vieses algorítmicos e racismo: o lado desconhecido da justiça algorítmica. Opinión Jurídica, Medellín, v. 23, n. 50, a49, p. 1-23, 2024. DOI: https://doi.org/10.22395/ojum.v23n50a49. Disponível em: https://revistas.udem.edu.co/index.php/opinion/pt_BR/article/view/4046. Acesso em: 1 set. 2026.

SAGAWA, Shiori et al. Distributionally robust neural networks. In: INTERNATIONAL CONFERENCE ON LEARNING REPRESENTATIONS, 8., 2020, [S. l.]. Proceedings [...]. [S. l.]: ICLR, 2020. Disponível em: https://openreview.net/pdf?id=ryxGuJrFvS. Acesso em: 31 out. 2026.

SCHLEDER, Gabriel R.; FAZZIO, Adalberto. Machine Learning na Física, Química, e Ciência de Materiais: descoberta e design de materiais. Revista Brasileira de Ensino de Física, v. 43, supl. 1, e20200407, 2021. DOI: https://doi.org/10.1590/1806-9126-RBEF-2020-0407. Disponível em: https://www.scielo.br/j/rbef/a/qzcfSKw4nzBK5Mddr8ZXy4v/?lang=pt. Acesso em: 1 set. 2026.

SCHROUFF, Jessica et al. Diagnosing failures of fairness transfer across distribution shift in real-world medical settings. In: ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS, 35., 2022, New Orleans. Proceedings [...]. Red Hook: Curran Associates, 2022. p. 19304-19318. Disponível em: https://proceedings.neurips.cc/paper_files/paper/2022/hash/7a969c30dc7e74d4e891c8ffb217cf79-Abstract.html. Acesso em: 31 out. 2026.

SCULLEY, D. et al. Hidden technical debt in machine learning systems. In: ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS, 28., 2015, Montreal. Proceedings [...]. Red Hook: Curran Associates, 2015. p. 2503-2511. Disponível em: https://papers.nips.cc/paper/5656-hidden-technical-debt-in-machine-learning-systems. Acesso em: 1 set. 2026.

SILVA, Flávia Helloysa Fontes et al. Inteligência artificial e desigualdade de gênero entre homens e mulheres: uma análise dos vieses em processos seletivos de contratação. Perspectivas em Gestão & Conhecimento, João Pessoa, v. 16, n. 1, p. 138-154, jan./abr. 2026. DOI: https://doi.org/10.22478/ufpb.2236-417X.2026v16n1.74614. Disponível em: https://periodicos.ufpb.br/index.php/pgc/article/view/74614. Acesso em: 1 set. 2026.

SINGH, Prerna. Systematic review of data-centric approaches in artificial intelligence and machine learning. Data Science and Management, v. 6, n. 3, p. 144-157, 2023. DOI: https://doi.org/10.1016/j.dsm.2023.06.001. Disponível em: https://www.sciencedirect.com/science/article/pii/S2666764923000279. Acesso em: 1 set. 2026.

SURESH, Harini; GUTTAG, John. A framework for understanding sources of harm throughout the machine learning life cycle. In: EQUITY AND ACCESS IN ALGORITHMS, MECHANISMS, AND OPTIMIZATION, 2021, [S. l.]. Proceedings [...]. New York: ACM, 2021. Disponível em: https://doi.org/10.1145/3465416.3483305. Acesso em: 31 out. 2026.

TRICCO, Andrea C. et al. PRISMA extension for scoping reviews (PRISMA-ScR): checklist and explanation. Annals of Internal Medicine, v. 169, n. 7, p. 467-473, 2018. Disponível em: https://doi.org/10.7326/M18-0850. Acesso em: 7 nov. 2026.

WHANG, Steven Euijong et al. Data collection and quality challenges in deep learning: a data-centric AI perspective. The VLDB Journal, v. 32, n. 4, p. 791-813, 2023. Disponível em: https://doi.org/10.1007/s00778-022-00775-9. Acesso em: 31 out. 2026.

YE, Wenqian et al. The Clever Hans mirage: a comprehensive survey on spurious correlations in machine learning. Transactions on Machine Learning Research, 2026. ISSN 2835-8856. Disponível em: https://openreview.net/forum?id=kIuqPmS1b1. Acesso em: 31 out. 2026.

Published

2026-11-19

How to Cite

Landim, C. E. M. P. ., & de Melo, M. J. A. . (2026). THE INFLUENCE OF DATA ENGINEERING ON DATA QUALITY FOR MACHINE LEARNING MODEL TRAINING: A NARRATIVE LITERATURE REVIEW. South American Sciences, 6(2), e26553. https://doi.org/10.63330/sasciencesv6n2-323

Most read articles by the same author(s)