A THREE-STAGE FAIRNESS-AWARE FRAMEWORK FOR RISK ASSESSMENT IN SOCIAL WORK: BALANCING ACCURACY AND EQUITY IN PREDICTIVE MODELS
Far East Journal of Electronics and Communications · 28.07.2026 · 30 (3) · 153-174
DOI: 10.17654/0973700626011 ↗Kurz gefasst
Fachartikel in der Zeitschrift Far East Journal of Electronics and Communications, online erschienen am 28.07.2026. Verfasst von Yih-Chang Chen und 2 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
The integration of Artificial Intelligence (AI) into high-stakes social work, such as child welfare and criminal justice, promises efficiency but risks perpetuating systemic biases. This study addresses the critical trade-off between predictive accuracy and algorithmic fairness. Baseline evaluations of Extreme Gradient Boosting (XGBoost) models across three datasets (COMPAS, AFST, and a synthetic dataset) revealed strong predictive performance (AUC up to 0.89) but significant racial bias. As shown in the fairness metrics comparison, pre-mitigation Statistical Parity Difference (SPD) values indicated severe bias (COMPAS: –0.21, AFST: –0.15, Synthetic: –0.18). To resolve this, we propose a “Three-Stage Fairness-Aware Framework” integrating pre-processing, in-processing, and post-processing mitigation strategies. The application of this framework successfully reduced bias to ethical thresholds, with post-mitigation SPD values improving significantly to –0.07, –0.04 and –0.06, respectively, while incurring minimal accuracy loss (< 5.1% AUC reduction). Furthermore, the study validates the framework’s computational scalability and its alignment with strict regulatory standards like the EU AI Act and GDPR. These findings provide an evidence-based blueprint for equitable and legally compliant AI governance in public services.
Zitieren
APA
Chen, Y., Lin, C., & Agan, S. (2026). A THREE-STAGE FAIRNESS-AWARE FRAMEWORK FOR RISK ASSESSMENT IN SOCIAL WORK: BALANCING ACCURACY AND EQUITY IN PREDICTIVE MODELS. Far East Journal of Electronics and Communications, 30(3), 153-174. https://doi.org/10.17654/0973700626011
Vancouver
Chen Y, Lin C, Agan S. A THREE-STAGE FAIRNESS-AWARE FRAMEWORK FOR RISK ASSESSMENT IN SOCIAL WORK: BALANCING ACCURACY AND EQUITY IN PREDICTIVE MODELS. Far East Journal of Electronics and Communications. 2026;30(3):153-174. doi:10.17654/0973700626011
BibTeX
@article{Chen2026,
author = {Chen, Yih-Chang and Lin, Chia-Ching and Agan, Sedat},
title = {A THREE-STAGE FAIRNESS-AWARE FRAMEWORK FOR RISK ASSESSMENT IN SOCIAL WORK: BALANCING ACCURACY AND EQUITY IN PREDICTIVE MODELS},
journal = {Far East Journal of Electronics and Communications},
year = {2026},
volume = {30},
number = {3},
pages = {153--174},
doi = {10.17654/0973700626011}
}
EUVIMED