Research Article

Customer Credit Risk Scoring Using Natural Language Processing (NLP): A News Analysis Approach

Volume: 7 Number: 2 December 8, 2025
EN TR

Customer Credit Risk Scoring Using Natural Language Processing (NLP): A News Analysis Approach

Abstract

The banking industry must make swift and reliable assessments in credit-granting processes that inherently involve high risk and must be grounded in information about the customer. In credit risk analysis, it is important to consider not only the customer’s past performance and financial position but also the news circulating in the media environment. However, it is not feasible to manually review the large volume of news produced on a daily basis. In this study, we aim to automatically process and score customer-related news by employing natural language processing and machine learning–based methods for automated sentiment analysis and named entity recognition. The system comprises the stages of collecting news from RSS sources, performing named entity recognition (NER) and sentiment analysis with Turkish BERT-based models, and then applying the scoring model. The developed framework enables the classification of news items as positive, negative, or neutral and allows these scores to be integrated into credit evaluation processes. Experimental results indicated that the proposed model reliably identifies and classifies the sentiment of customer-related news items in a manner consistent with business experts’ assessments. The evaluation confirmed that the system can automatically collect and analyze news content, supporting faster and more objective decision-making compared to manual review processes.

Keywords

References

  1. [1] SU, J., LAU, R. Y. K., YU, J., NG, D. C. T. ve JIANG, W. (2025). A multi-modal data fusion approach for evaluating the impact of extreme public sentiments on corporate credit ratings. Complex & Intelligent Systems, 11, 436. https://doi.org/10.1007/s40747-025-02067-5
  2. [2] Savasy. (2020). Turkish BERT models for NLP tasks [Computer software]. Hugging Face. https://huggingface.co/savasy
  3. [3] DU, K., ZHAO, Y., MAO, R., XING, F. ve CAMBRIA, E. (2024). Natural Language Processing in Finance: A Survey. Information Fusion, 115, 102755. https://doi.org/10.1016/j.inffus.2024.102755
  4. [4] DEVEIKYTE, J., GEMAN, H., PICCARI, C. ve PROVETTI, A. (2020). A Sentiment Analysis Approach to the Prediction of Market Volatility. arXiv preprint arXiv:2012.05906. https://arxiv.org/abs/2012.05906
  5. [5] How do machine learning and non-traditional data affect credit scoring? New evidence from a Chinese fintech firm (2024). Journal of Financial Stability, 73, 101284. https://doi.org/10.1016/j.jfs.2024.101284
  6. [6] Aydin, C. R. and Gungor, T. (2021). Sentiment analysis in Turkish: Supervised, semi-supervised, and unsupervised techniques. NATURAL LANGUAGE ENGINEERING, 27(4), 455–483. https://doi.org/10.1017/S1351324920000200
  7. [7] GÖKTÜRK, İ. E. and Soydan, E. (2021). Zihinsel muhasebe eğilimi çerçevesinde bireylerde finansal karar alma süreci: finansal sosyal hizmet üzerine bir değerlendirme. Toplum Ve Sosyal Hizmet, 32(4), 1543-1564. https://doi.org/10.33417/tsh.903821
  8. [8] Çetin, F. S. and Eryiğit, G. (2018). Türkçe hedef tabanlı duygu analizi i̇çin alt görevlerin i̇ncelenmesi – hedef terim, hedef kategori ve duygu sınıfı belirleme. Bilişim Teknolojileri Dergisi, 11(1), 43-56. https://doi.org/10.17671/gazibtd.325865

Details

Primary Language

English

Subjects

Natural Language Processing

Journal Section

Research Article

Early Pub Date

December 3, 2025

Publication Date

December 8, 2025

Submission Date

October 31, 2025

Acceptance Date

November 25, 2025

Published in Issue

Year 2025 Volume: 7 Number: 2

APA
Sarı, H., & Turan, M. (2025). Customer Credit Risk Scoring Using Natural Language Processing (NLP): A News Analysis Approach. International Journal of Engineering and Innovative Research, 7(2), 138-152. https://doi.org/10.47933/ijeir.1814441
AMA
1.Sarı H, Turan M. Customer Credit Risk Scoring Using Natural Language Processing (NLP): A News Analysis Approach. IJEIR. 2025;7(2):138-152. doi:10.47933/ijeir.1814441
Chicago
Sarı, Hayri, and Metin Turan. 2025. “Customer Credit Risk Scoring Using Natural Language Processing (NLP): A News Analysis Approach”. International Journal of Engineering and Innovative Research 7 (2): 138-52. https://doi.org/10.47933/ijeir.1814441.
EndNote
Sarı H, Turan M (December 1, 2025) Customer Credit Risk Scoring Using Natural Language Processing (NLP): A News Analysis Approach. International Journal of Engineering and Innovative Research 7 2 138–152.
IEEE
[1]H. Sarı and M. Turan, “Customer Credit Risk Scoring Using Natural Language Processing (NLP): A News Analysis Approach”, IJEIR, vol. 7, no. 2, pp. 138–152, Dec. 2025, doi: 10.47933/ijeir.1814441.
ISNAD
Sarı, Hayri - Turan, Metin. “Customer Credit Risk Scoring Using Natural Language Processing (NLP): A News Analysis Approach”. International Journal of Engineering and Innovative Research 7/2 (December 1, 2025): 138-152. https://doi.org/10.47933/ijeir.1814441.
JAMA
1.Sarı H, Turan M. Customer Credit Risk Scoring Using Natural Language Processing (NLP): A News Analysis Approach. IJEIR. 2025;7:138–152.
MLA
Sarı, Hayri, and Metin Turan. “Customer Credit Risk Scoring Using Natural Language Processing (NLP): A News Analysis Approach”. International Journal of Engineering and Innovative Research, vol. 7, no. 2, Dec. 2025, pp. 138-52, doi:10.47933/ijeir.1814441.
Vancouver
1.Hayri Sarı, Metin Turan. Customer Credit Risk Scoring Using Natural Language Processing (NLP): A News Analysis Approach. IJEIR. 2025 Dec. 1;7(2):138-52. doi:10.47933/ijeir.1814441

88x31.png

This work is licensed under a Creative Commons Attribution 4.0 International License