AI-Powered Resume Screening: Opportunities, Biases, and Interpretability Challenges
DOI:
https://doi.org/10.21590/ehz64g49Abstract
Automated resume screening tools are increasingly adopted by HR departments to streamline talent acquisition. However, concerns around algorithmic bias, fairness, and explainability have drawn scrutiny from regulators and researchers alike. This paper investigates the use of AI models—particularly natural language processing (NLP) and machine learning (ML)—in resume parsing, ranking, and filtering. We develop and evaluate three models: a logistic regression baseline with TF-IDF features, a fine-tuned BERT model for semantic understanding, and a gradient-boosted decision tree (XGBoost) trained on hand-labeled hiring outcomes from a publicly available dataset. BERT-based models improve prediction accuracy by 15% over traditional keyword matching, identifying relevant experience even with unstructured or unconventional formats. However, interpretability suffers due to the opaque nature of deep language models. We analyze bias using gender-swapped resumes and show that both BERT and XGBoost models exhibit measurable disparities in ranking outcomes, favoring traditionally male-coded language and experience gaps. Feature importance and SHAP value visualizations are used to probe decision logic. Our study highlights the tension between performance and fairness in AI-based hiring tools. We propose a hybrid approach combining interpretable shallow models for screening with deep models for contextual scoring, alongside human-in-the-loop validation. This paper contributes guidelines for responsible AI use in recruitment systems.
References
1. Bolukbasi, T., Chang, K. W., Zou, J. Y., Saligrama, V., & Kalai, A. T. (2016). Man is to computer programmer as woman is to homemaker? Debiasing word embeddings. Proceedings of the 30th International Conference on Neural Information Processing Systems (NeurIPS), 4349 4357.
2. Talluri Durvasulu, M. B. (2017). AWS Storage: Key Concepts for Solution Architects. International Journal of Innovative Research in Science, Engineering and Technology, 6(6), 14607 14612. https://doi.org/10.15680/IJIRSET.2017.0606352
3. Ribeiro, M. T., Singh, S., & Guestrin, C. (2016). “Why should I trust you?”: Explaining the predictions of any classifier. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 1135 1144.
4. Lundberg, S. M., & Lee, S. I. (2017). A unified approach to interpreting model predictions. Advances in Neural Information Processing Systems (NeurIPS), 4765
4774.
5. Devlin, J., Chang, M. W., Lee, K., & Toutanova, K. (2019). BERT: Pre training of deep bidirectional transformers for language understanding. Proceedings of NAACL HLT, 4171 4186.
6. Bellamkonda, S. (2016). Network Switches Demystified: Boosting Performance and Scalability. NeuroQuantology, 14(1), 193 196.
7. Liem, C. C. S., Langer, M., Demetriou, A., Liu, J., & Sörensen, J. (2018). Psychology meets machine learning: Interdisciplinary perspectives on algorithmic job candidate screening. Proceedings of the 2018 AAAI/ACM Conference on AI, Ethics, and Society (AIES), 306 312.
8. Binns, R., Veale, M., Van Kleek, M., & Shadbolt, N. (2018). ‘It’s reducing a human being to a percentage’: Perceptions of justice in algorithmic decisions. Proceedings of the 2018 CHI Conference on Human Factors in Computing Systems, 1 14.
9. Guo, C., & Berkhahn, F. (2016). Entity embeddings of categorical variables. arXiv preprint arXiv:1604.06737.
10. Kolla, S. (2018). Legacy liberation: Transitioning to cloud databases for enhanced agility and innovation. International Journal of Computer Engineering and Technology, 9(2), 237 248. https://doi.org/10.34218/IJCET_09_02_023
11. Tolan, S., Miron, M., Gomez, E., & Castillo, C. (2019). Why machine learning may lead to unfairness: Evidence from risk prediction in criminal sentencing. Proceedings of the ACM Conference on Fairness, Accountability, and Transparency, 230 239.
12. Barocas, S., Hardt, M., & Narayanan, A. (2019). Fairness and machine learning. fairmlbook.org. 13. Kamiran, F., & Calders, T. (2012). Data preprocessing techniques for classification without discrimination. Knowledge and Information Systems, 33(1), 1 33.
14. Feldman, M., Friedler, S. A., Moeller, J., Scheidegger, C., & Venkatasubramanian, S. (2015). Certifying and removing disparate impact. Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 259
268.
15. Satish Kumar Nalluri, Venkata Krishna Bharadwaj Parasaram. (2019). Software Centric Automation Frameworks Integrating AI and Cybersecurity Principles. International Journal of Engineering Science & Humanities, 9(1), 30 40. Retrieved from https://www.ijesh.com/j/article/view/539
16. Holstein, K., Wortman Vaughan, J., Daumé III, H., Dudik, M., & Wallach, H. (2019). Improving fairness in machine learning systems: What do industry practitioners need? Proceedings of the 2019 CHI Conference on Human Factors in Computing Systems, 1 16.
17. Goli, V. R. (2018). Optimizing and Scaling Large Scale Angular Applications: Performance, Side Effects, Data Flow, and Testing. International Journal of Innovative Research in Science, Engineering and Technology, 7(2), 1181 1184. https://www.ijirset.com/upload/2018/february/1_Optimizing1.pdf
18. Angwin, J., Larson, J., Mattu, S., & Kirchner, L. (2016). Machine bias. ProPublica. Retrieved from https://www.propublica.org/article/machine bias risk assessments
in criminal sentencing
19. Jobin, A., Ienca, M., & Vayena, E. (2019). The global landscape of AI ethics guidelines. Nature Machine Intelligence, 1(9), 389 399.
