Fairness Evaluation in Text Classification: Machine Learning Practitioner Perspectives of Individual and Group Fairness

Abstract

Mitigating algorithmic bias is a critical task in the development and deployment of machine learning models. While several toolkits exist to aid machine learning practitioners in addressing fairness issues, little is known about the strategies practitioners employ to evaluate model fairness and what factors influence their assessment, particularly in the context of text classification. Two common approaches of evaluating the fairness of a model are group fairness and individual fairness. We run a study with Machine Learning practitioners (n=24) to understand the strategies used to evaluate models. Metrics presented to practitioners (group vs. individual fairness) impact which models they consider fair. Participants focused on risks associated with underpredicting / overpredicting and model sensitivity relative to identity token manipulations. We discover fairness assessment strategies involving personal experiences or how users form groups of identity tokens to test model fairness. We provide recommendations for interactive tools for evaluating fairness in text classification. CCS CONCEPTS• Human-centered computing → Human computer interaction (HCI).

Data Ethics Emergency Drill: A Toolbox for Discussing Responsible AI for Industry Teams

“…A number of studies addressed the question of what practitioners need from responsible AI by asking them directly. ML practitioners call for context-specific tools [28] or ways of making contextual information evident in existing tools [3]. They ask for tools that fit into their resource constraints, as ML practices outside of big tech may not have the bandwidth to carry out some of the responsible AI investigations of larger companies [29].…”