Toward a Pipeline-aware View of Fairness for Machine Learning

While algorithmic fairness continues to be a thriving area of research in ML, in practice, mitigating issues of bias and unfairness often gets reduced to enforcing an arbitrarily chosen fairness metric, either by enforcing fairness constraints during the optimization step, post-processing model outputs or by manipulating the training data. Recent work has called on the ML community to take a more holistic approach to tackle fairness issues by systematically investigating the multitude of design choices made through the ML pipeline and identifying effective interventions at the root cause, as opposed to the symptoms. However, there are currently very few methods of operationalizing this approach in practice. Drawing on our experience as educators and practitioners, we first demonstrate that without clear guidelines and toolkits, even individuals with specialized ML knowledge find it challenging to hypothesize how various design choices influence model behavior. We then consult the fair-ML literature to understand the progress to date toward operationalizing the pipeline-aware approach: we systematically collect and organize the prior work that attempts to detect, measure, and mitigate various sources of unfairness through the ML pipeline. We utilize this extensive categorization of previous contributions to sketch a research agenda for the community.