bias mitigation algorithm design and implementation

The widespread deployment of machine learning (ML) in decision-making systems introduces significant fairness risks—particularly concerning the handling of sensitive attributes and the protection of minority groups—while software engineering lacks a systematic, lifecycle-oriented framework for fairness engineering practices. Method: We conduct a systematic mapping study (SMS) combined with a comprehensive literature review to analyze fairness-related practices across the ML development lifecycle. Contribution/Results: We propose the first software engineering–centric fairness practice taxonomy, comprising 28 structured, actionable practices explicitly mapped to data preprocessing, modeling, and deployment stages. Each practice is annotated with its corresponding ML lifecycle phase and contextual applicability, thereby bridging the gap between fairness research and industrial implementation. This taxonomy serves as an integrable, operational guide for researchers and practitioners, enhancing the reliability, accountability, and trustworthiness of ML systems.

This paper addresses selection and label bias in fairness evaluation for automated decision-making (e.g., hiring), which leads mainstream interventions—such as resampling—to produce a “fairness illusion.” To mitigate this, we propose an audit-inspired fictitious applicant experimental framework integrating randomized controlled trials, base-rate parity, and individual treatment effect (ITE) estimation. Our method enables more realistic fairness assessment and model training. Key contributions include: (i) the first incorporation of ITE estimation into algorithmic fairness intervention, thereby reducing overreliance on aggregate group-level metrics; and (ii) empirical evidence demonstrating that conventional methods conceal approximately 10% residual discrimination, whereas our approach significantly reduces actual discriminatory outcomes. The framework enhances internal validity of fairness evaluation and improves intervention efficacy.

Post-processing debiasing methods may inadvertently introduce new forms of unfairness—particularly through overcorrection caused by imbalanced prediction flips across demographic groups. To address this, we propose “Flip Disparity,” a novel metric suite that quantifies, for the first time in post-processing, the relative proportion of predictions flipped per group, thereby overcoming limitations of conventional fairness metrics that ignore transparency and proportionality in correction behavior. Our method leverages differences in confusion matrices and inter-group comparative analysis, integrated within a unified framework combining visual diagnostic tools and strategy-comparability assessment. This paradigm significantly enhances the interpretability of debiasing strategies and enables reliable detection of latent imbalanced corrections. Empirical evaluation across multiple benchmark datasets reveals previously undetected correction biases in widely adopted fairness algorithms. The proposed framework establishes a verifiable, auditable standard for responsible algorithmic governance.

Low adoption of machine learning fairness intervention tools stems primarily from developers’ limited awareness and insufficient evidence for comparative evaluation. This study presents the first large-scale empirical survey of 62 open-source fairness tools, integrating systematic literature analysis, codebase mining, maintainability assessment, and a functional classification framework to characterize their technical attributes (e.g., in-processing as the dominant intervention stage), evolutionary trajectories, and practical utility. Results show that 50% of tools support both bias detection and mitigation; only 32% have sustained active maintenance within the past year; and over half provide end-to-end workflow support. The analysis uncovers a critical tension between developer preferences—favoring integrated, usable tooling—and tool sustainability, revealing significant gaps in long-term maintenance and documentation. These findings establish a foundational empirical basis for improving fairness tool design, guiding practitioner selection, and informing ecosystem development strategies.

To address the challenge of sustaining fairness auditing for deployed models under data policy drift, subgroup heterogeneity, and dynamic distributional shifts, this paper proposes the first dynamic fairness auditing framework tailored for streaming data. Methodologically, it innovatively adapts “testing by betting”—a game-theoretic statistical paradigm—into fairness monitoring, integrating anytime-valid inference with nonparametric, sample-size-agnostic sequential testing that adapts to time-varying sampling strategies. Unlike conventional batch-based approaches, our framework enables real-time, continuous fairness assessment without requiring prior knowledge of data distribution or fixed sample sizes. Experiments on three benchmark datasets demonstrate that the method achieves high detection sensitivity, low false positive rate, and strong robustness against concept drift and subgroup imbalance. These properties collectively enhance the practicality, timeliness, and deployability of fairness monitoring in real-world operational environments.

This study addresses the lack of systematic evaluation of robustness in existing fair machine learning methods under realistic data perturbations such as label noise, missing data, and distribution shifts. It introduces a causal inference framework to conduct the first comprehensive robustness analysis of mainstream fairness interventions—including sensitive attribute handling and bias mitigation techniques—under non-ideal data conditions. Empirical results demonstrate that several widely used approaches suffer significant performance degradation under common perturbations, thereby exposing critical limitations for real-world deployment. These findings provide both theoretical grounding and practical guidance for developing more reliable and robust fair machine learning systems.

This study addresses fairness deficiencies in machine learning–based early warning systems used by higher education institutions for allocating student support resources, particularly with respect to disparities arising from gender, age, and residency status. Through a long-term collaboration with Centennial College, the authors replicate the institution’s deployed system and develop the first reproducible auditing framework that integrates construct validity with statistical fairness metrics to systematically evaluate the entire pipeline—from data collection and prediction to post-processing. Their analysis reveals that younger, male, and international students are systematically assigned higher risk scores than their actual risk levels warrant, while older and female students with equivalent risk profiles are consistently underestimated. Notably, bias is significantly amplified during the post-processing stage. This work provides both methodological innovation and empirical evidence to advance fairness auditing of institutionalized machine learning systems.

This work addresses the challenge of auditing external fairness in real-world AI systems, where auditors typically have only limited query access to model outputs. Existing approaches rely on static datasets and are ill-suited for sequential auditing scenarios. To bridge this gap, the paper introduces the first sequential fairness auditing framework tailored to practical deployment constraints, formulating the audit as a tolerance-aware sequential hypothesis test. Built upon generalized likelihood ratio methods, the framework accommodates diverse model outputs—including decisions, scores, and logits—and dynamically accumulates evidence, enabling early termination once a conclusive determination is reached. Empirical results demonstrate that both the type of fairness metric and the richness of model output significantly influence auditing efficiency: richer outputs generally reduce query complexity substantially, though their advantage diminishes when the system’s fairness level is near the decision threshold.

This work addresses the persistent underperformance of machine learning models on intersecting sensitive subgroups—such as those defined by race and gender—attributed to inadequate bias metrics and insufficient representation in training data. The authors introduce coverage constraints into a bias mitigation framework for the first time, formulating the problem via integer linear programming to optimize data modification costs while ensuring adequate representation across all (intersecting) groups and bounding approximation error in bias reduction. This approach enables quantification of the “price of fairness,” facilitating principled trade-offs between equity and data efficiency in legal compliance and data governance contexts. Empirical evaluations across multiple benchmark datasets and classifiers demonstrate that the proposed framework effectively preserves predictive accuracy while substantiating the critical role of coverage constraints in safeguarding both fairness and performance in downstream models.

This study addresses the limitations of prevailing algorithmic fairness assessments, which often focus narrowly on technical metrics while neglecting organizational and societal contexts, thereby failing to uncover systemic biases in real-world deployments. It presents the first end-to-end socio-technical audit of a semi-automated hiring system used by Barcelona’s public employment service from 2017 to 2022, analyzing nearly 500,000 candidate–job pipeline records. Integrating disparate impact ratio (DIR), multi-stage tracking, and intersectional fairness measures across gender, age, and salary levels, the analysis reveals that while overall gender representation appears balanced, women are significantly underrepresented in shortlists for mid-salary positions (DIR = 0.786), non-binary individuals are selected at less than one-third the rate of men, and candidates over 55 are entirely absent. The audit further uncovers process-level biases invisible to model-centric evaluations and highlights critical information asymmetries between vendors and deploying institutions.