Abstract
We investigate the fairness of local explanations in AI models by comparing the mean explanations for privileged and unprivileged groups across various datasets and model types. Specifically, we train linear, multi-layer perceptron, and explainable boosting machine (EBM) models, on several key academic datasets. Local explanations are generated using both post-hoc methods (LIME), and direct methods - logistic regression, integrated gradients and EBM local feature importance. By comparing these explanations across models and methods, we introduce a new metric, the Local Explanation Amplification Disparity Ratio (LEADR), to measure disparities in feature attribution between privileged and unprivileged groups. Our preliminary findings suggest that transparent Whitebox models may exhibit a tendency to display greater disparity in bias than opaque Blackbox models. This insight encourages further research into bias mitigation strategies that are tailored to different algorithm types, with the goal of minimizing undesired bias in AI systems.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
References
Chhillar, D., Aguilera, R.V.: An eye for artificial intelligence: insights into the governance of artificial intelligence and vision for future research. Bus. Soc. 61(5), 1197–1241 (2022)
Chizari, N., Shoeibi, N., Moreno-García, M.N.: A comparative analysis of bias amplification in graph neural network approaches for recommender systems. Electronics 11(20) 3301 (2022)
Lu, Q., et al.: Responsible-AI-by-design: a pattern collection for designing responsible AI systems. IEEE Softw. (2023)
Subramanian, A., et al.: Spatial-frequency channels, shape bias, and adversarial robustness. In: Advances in Neural Information Processing Systems, vol. 36 (2024)
Quy, T.L., et al.: A survey on datasets for fairness-aware machine learning. Wiley Interdisc. Rev. Data Min. Knowl. Discov. 12(3), e1452 (2022)
Ehsan, U., et al.: The who in explainable AI: how AI background shapes perceptions of AI explanations. arXiv preprint arXiv:2107.13509 (2021)
A novel explainable AI-based approach to estimate the natural period of vibration of masonry infill reinforced concrete frame structures using different machine learning techniques. Results Eng. 19, 101388 (2023). https://doi.org/10.1016/j.rineng2023.101388. https://www.sciencedirect.com/science/article/pii/S2590123023005157. ISSN 2590-1230
Vale, D., El-Sharif, A., Ali, M.: Explainable artificial intelligence (XAI) post-hoc explainability methods: risks and limitations in non-discrimination law. AI Ethics 2(4), 815–826 (2022)
Dai, J., et al.: Fairness via explanation quality: evaluating disparities in the quality of post hoc explanations. In: Proceedings of the 2022 AAAI/ACM Conference on AI, Ethics, and Society, pp. 203–214 (2022)
Chang, P.W., Fishman, L., Neel, S.: Model Explanation Disparities as a Fairness Diagnostic (2023)
Kim, D., et al.: How should the results of artificial intelligence be explained to users?-research on consumer preferences in user-centered explainable artificial intelligence. Technol. Forecast. Soc. Change 188, 122343 (2023)
Kohavi, R., et al.: Scaling up the accuracy of Naive-Bayes classifiers: a decision-tree hybrid. In: KDD, vol. 96, pp. 202–207 (1996)
Akhavan Rahnama, A.H.: The blame problem in evaluating local explanations, and how to tackle it. arXiv preprint arXiv:2310.03466 (2023)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Ethics declarations
Disclosure of Interests
The views reflected in this article are the views of the authors and do not necessarily reflect the views of the global EY organization or its member firms.
Rights and permissions
Copyright information
© 2025 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Shoeibi, N., DeGange, J., Shoeibi, N., Shoeibi, A. (2025). Measuring Fairness in AI Explanations with LEADR: Local Explanation Amplification Disparity Ratio. In: Mathieu, P., De la Prieta, F. (eds) Advances in Practical Applications of Agents, Multi-Agent Systems, and Digital Twins: The PAAMS Collection. PAAMS 2024. Lecture Notes in Computer Science(), vol 15157. Springer, Cham. https://doi.org/10.1007/978-3-031-70415-4_22
Download citation
DOI: https://doi.org/10.1007/978-3-031-70415-4_22
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-70414-7
Online ISBN: 978-3-031-70415-4
eBook Packages: Computer ScienceComputer Science (R0)