SavedAdd your journal

Journals / Black Sea Journal of Engineering and Science / 2026 / Cilt: 9 Sayı: 4

Correcting Erroneous Entries in Datasets Using Fuzzy Matching Methods and Its Impact on Machine Learning Performance

Bulanık Eşleşme Yöntemi ile Veri Setlerindeki Hatalı Girişlerin Düzeltilmesi ve Makine Öğrenmesi Performansına Etkisi

Ahmet Tezcan Tekin(İSTANBUL TEKNİK ÜNİVERSİTESİ, TEKNİK OKUL)0000-0002-1792-6622
Pages
2055–2064
Publication Date
2026-07-15
🏛️Bibliographic metadata and full text access for this article are provided via DergiPark.
DergiPark Record ↗
✨

AI Summary & Key Takeaways

AI Insight
⚡ TL;DR(Too Long; Didn't Read)

Experimental results demonstrated that fuzzy matching-based data correction improved the F1-score by an average of +8.9 percentage points across all models and that the corrected dataset produced performance very close to the original error-free data.

📌 3 Key Takeaways

🎯
Objective & Hypothesis:Data quality is one of the most critical factors directly affecting the performance of machine learning models.
📊
Key Findings & Data:In this study, spelling errors, case inconsistencies, abbreviation variations, and transposition errors were injected in a controlled manner into a synthetic customer segmentation dataset consisting of 10,000 records.
💡
Conclusion & Impact:Experimental results demonstrated that fuzzy matching-based data correction improved the F1-score by an average of +8.9 percentage points across all models and that the corrected dataset produced performance very close to the original error-free data.
🔬Design:Modeling & Simulation
📍Scope:Comprehensive Sample
⚙️Variables:Bulanık eşleşme · Veri kalitesi · Veri temizleme

Abstract

Data quality is one of the most critical factors directly affecting the performance of machine learning models. Spelling errors, inconsistent encodings, and non-standard entries are frequently encountered in real-world datasets and adversely affect the model training process. This study examines the role of fuzzy matching methods in detecting and correcting erroneous entries during the data preprocessing stage. The mathematical foundations of fundamental fuzzy matching algorithms, including Levenshtein distance, Jaro-Winkler similarity, and n-gram-based approaches, are systematically presented, and how these methods can be used to resolve inconsistencies in categorical variables is discussed. In this study, spelling errors, case inconsistencies, abbreviation variations, and transposition errors were injected in a controlled manner into a synthetic customer segmentation dataset consisting of 10,000 records. A three-stage correction pipeline based on Jaro-Winkler similarity was applied, successfully correcting 96.3% of the erroneous entries. The performances of classification algorithms such as Logistic Regression, Decision Tree, Random Forest, and XGBoost were comparatively evaluated on both corrected and uncorrected datasets. Experimental results demonstrated that fuzzy matching-based data correction improved the F1-score by an average of +8.9 percentage points across all models and that the corrected dataset produced performance very close to the original error-free data. Furthermore, the correction process was observed to significantly improve computational efficiency by reducing the feature space by 38% after one-hot encoding.

Özet

Veri kalitesi, makine öğrenmesi modellerinin başarısını doğrudan etkileyen en kritik faktörlerden biridir. Gerçek dünya veri setlerinde sıklıkla karşılaşılan yazım hataları, tutarsız kodlamalar ve standart dışı girişler, model eğitim sürecini olumsuz yönde etkilemektedir. Bu çalışma, bulanık eşleşme (fuzzy matching) yöntemlerinin veri ön işleme aşamasında hatalı girişlerin tespiti ve düzeltilmesindeki rolünü incelemektedir. Levenshtein mesafesi, Jaro-Winkler benzerliği ve n-gram tabanlı yaklaşımlar gibi temel bulanık eşleşme algoritmalarının matematiksel temelleri sistematik biçimde ele alınmış ve bu yöntemlerin kategorik değişkenlerdeki tutarsızlıkları gidermede nasıl kullanılabileceği tartışılmıştır. Çalışmada, 10.000 kayıttan oluşan sentetik bir müşteri segmentasyonu veri setine kontrollü biçimde yazım hataları, büyük/küçük harf tutarsızlıkları, kısaltma farklılıkları ve transpozisyon hataları enjekte edilmiştir. Jaro-Winkler benzerliğine dayalı üç aşamalı bir düzeltme pipeline'ı uygulanarak hatalı girişlerin %96,3'ü başarılı biçimde düzeltilmiştir. Düzeltilmiş ve düzeltilmemiş veri setleri üzerinde Lojistik Regresyon, Karar Ağacı, Rastgele Orman ve XGBoost gibi sınıflandırma algoritmalarının performansları karşılaştırmalı olarak değerlendirilmiştir. Deneysel sonuçlar, bulanık eşleşme tabanlı veri düzeltmenin tüm modellerde F1 skorunu ortalama +8,9 yüzde puan artırdığını ve düzeltilmiş veri setinin orijinal hatasız veriye çok yakın performans ürettiğini ortaya koymuştur. Ayrıca düzeltme işleminin one-hot encoding sonrası öznitelik uzayını %38 oranında daraltarak hesaplama verimliliğini de önemli ölçüde iyileştirdiği gözlemlenmiştir.

🔖

Cite This Article

Tekin, A. T. (2026). Correcting Erroneous Entries in Datasets Using Fuzzy Matching Methods and Its Impact on Machine Learning Performance. Black Sea Journal of Engineering and Science, Cilt: 9 Sayı: 4, 2055–2064. https://doi.org/10.34248/bsengineering.1903764

📄 Full Text PDF Viewer

📄

PDF görüntüleyici hazırlanıyor...

References (25)

  1. Bilenko, M., & Mooney, R. J. (2003). Adaptive duplicate detection using learnable string similarity measures. In Proceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (pp. 39–48). ACM.
  2. Christen, P. (2012). Data matching: Concepts and techniques for record linkage, entity resolution, and duplicate detection. Springer.
  3. Cohen, W. W., Ravikumar, P., & Fienberg, S. E. (2003). A comparison of string distance metrics for name-matching tasks. In Proceedings of the IJCAI Workshop on Information Integration on the Web (pp. 73–78).
  4. CrowdFlower. (2016). 2016 data science report. CrowdFlower Inc. https://web.archive.org/web/20180801063101/http://visit.crowdflower.com/rs/416-ZBE-142/images/CrowdFlower_DataScienceReport_2016.pdf (Erişim tarihi: 13 Temmuz 2026)
  5. Damerau, F. J. (1964). A technique for computer detection and correction of spelling errors. Communications of the ACM, 7(3), 171–176.
  6. Elmagarmid, A. K., Ipeirotis, P. G., & Verykios, V. S. (2007). Duplicate record detection: A survey. IEEE Transactions on Knowledge and Data Engineering, 19(1), 1–16.
  7. Frénay, B., & Verleysen, M. (2014). Classification in the presence of label noise: A survey. IEEE Transactions on Neural Networks and Learning Systems, 25(5), 845–869.
  8. Hameed, M., & Naumann, F. (2020). Data preparation: A survey of commercial tools. ACM SIGMOD Record, 49(3), 18–29.
  9. Hancock, J. T., & Khoshgoftaar, T. M. (2020). Survey on categorical data for neural networks. Journal of Big Data, 7, Article 28.
  10. Ilyas, I. F., & Chu, X. (2019). Data cleaning. Morgan & Claypool Publishers.
  11. Jaro, M. A. (1989). Advances in record-linkage methodology as applied to matching the 1985 census of Tampa, Florida. Journal of the American Statistical Association, 84(406), 414–420.
  12. Kilkenny, M. F., & Robinson, K. M. (2018). Data quality: "Garbage in – garbage out." Health Information Management Journal, 47(3), 103–105.
  13. Kondrak, G. (2005). N-gram similarity and distance. In Proceedings of the International Symposium on String Processing and Information Retrieval (SPIRE), LNCS 3772 (pp. 115–126). Springer.
  14. Krishnan, S., Wang, J., Wu, E., Franklin, M. J., & Goldberg, K. (2016). ActiveClean: Interactive data cleaning for statistical modeling. Proceedings of the VLDB Endowment, 9(12), 948–959.
  15. Levenshtein, V. I. (1966). Binary codes capable of correcting deletions, insertions, and reversals. Soviet Physics Doklady, 10(8), 707–710.
  16. Li, P., Dong, X. L., Maurino, A., & Srivastava, D. (2011). Linking temporal records. Proceedings of the VLDB Endowment, 4(11), 956–967.
  17. Li, Y., Li, J., Suhara, Y., Doan, A., & Tan, W.-C. (2020). Deep entity matching with pre-trained language models. Proceedings of the VLDB Endowment, 14(1), 50–60.
  18. Mahdavi, M., & Abedjan, Z. (2020). Baran: Effective error correction via a unified context representation and transfer learning. Proceedings of the VLDB Endowment, 13(12), 1948–1961.
  19. Mudgal, S., Li, H., Rekatsinas, T., Doan, A., Park, Y., Krishnan, G., Deep, R., Arcaute, E., & Raghavendra, V. (2018). Deep learning for entity matching: A design space exploration. In Proceedings of the ACM SIGMOD International Conference on Management of Data (pp. 19–34). ACM.
  20. Müller, H., & Freytag, J. C. (2003). Problems, methods, and challenges in comprehensive data cleansing (Technical Report HUB-IB-164). Humboldt-Universität zu Berlin.
  21. Nettleton, D. F., Orriols-Puig, A., & Fornells, A. (2010). A study of the effect of different types of noise on the precision of supervised learning techniques. Artificial Intelligence Review, 33(4), 275–306.
  22. Rahm, E., & Do, H. H. (2000). Data cleaning: Problems and current approaches. IEEE Data Engineering Bulletin, 23(4), 3–13.
  23. Sokolova, M., & Lapalme, G. (2009). A systematic analysis of performance measures for classification tasks. Information Processing & Management, 45(4), 427–437.
  24. Winkler, W. E. (1990). String comparator metrics and enhanced decision rules in the Fellegi-Sunter model of record linkage. In Proceedings of the Section on Survey Research Methods (pp. 354–359). American Statistical Association.
  25. Zhu, X., & Wu, X. (2004). Class noise vs. attribute noise: A quantitative study. Artificial Intelligence Review, 22(3), 177–210.