FairSense: Long-Term Fairness Analysis of ML-Enabled Systems

Algorithmic fairness of machine learning (ML) models has raised significant concern in the recent years. Many testing, verification, and bias mitigation techniques have been proposed to identify and reduce fairness issues in ML models. The existing methods are model-centric and designed to detect fa...

Ausführliche Beschreibung

Gespeichert in:

Bibliographische Detailangaben
Veröffentlicht in:	Proceedings / International Conference on Software Engineering S. 782 - 794
Hauptverfasser:	She, Yining, Biswas, Sumon, Kastner, Christian, Kang, Eunsuk
Format:	Tagungsbericht
Sprache:	Englisch
Veröffentlicht:	IEEE 26.04.2025
Schlagworte:	Environmental factors Feedback loop long-term fairness Machine learning algorithms ml-enabled system Monte Carlo methods Opioids Prediction algorithms Prevention and mitigation Sensitivity analysis simulation-based analysis Software engineering system design Testing
ISSN:	1558-1225
Online-Zugang:	Volltext
Tags:	Tag hinzufügen Keine Tags, Fügen Sie den ersten Tag hinzu!

Beschreibung
Zusammenfassung:	Algorithmic fairness of machine learning (ML) models has raised significant concern in the recent years. Many testing, verification, and bias mitigation techniques have been proposed to identify and reduce fairness issues in ML models. The existing methods are model-centric and designed to detect fairness issues under static settings. However, many ML-enabled systems operate in a dynamic environment where the predictive decisions made by the system impact the environment, which in turn affects future decision-making. Such a self-reinforcing feedback loop can cause fairness violations in the long term, even if the immediate outcomes are fair. In this paper, we propose a simulation-based framework called Fairsenseto detect and analyze long-term unfairness in ML-enabled systems. Given a fairness requirement, Fairsenseperforms Monte-Carlo simulation to enumerate evolution traces for each system configuration. Then, Fairsenseperforms sensitivity analysis on the space of possible configurations to understand the impact of design options and environmental factors on the long-term fairness of the system. We demonstrate Fairsense'spotential utility through three real-world case studies: Loan lending, opioids risk scoring, and predictive policing.
ISSN:	1558-1225
DOI:	10.1109/ICSE55347.2025.00159