A Catalog of Fairness-Aware Practices in Machine Learning Engineering

Abstract

Artificial Intelligence (AI)’s widespread adoption in decision-making processes, particularly with the introduction of AI-based assistants, raises concerns about ethics and fairness, particularly regarding the treatment of sensitive features and potential discrimination against underrepresented groups. The software engineering community has responded by developing fairness-oriented metrics, empirical studies, and mitigation approaches. However, there remains a gap in understanding and categorizing practices for engineering fairness throughout the development lifecycle of AI-based solutions. This paper presents a catalog of practices for addressing fairness derived from a systematic mapping study. The study identifies and categorizes 28 practices from existing literature, mapping them onto different stages of the development lifecycle. From this catalog, actionable items and implications for both researchers and practitioners in software engineering were extracted. This work aims to provide a comprehensive resource for integrating fairness considerations into the development and deployment of AI systems, enhancing their reliability, accountability, and credibility.

Ähnliche Arbeiten