Transferring Fairness using Multi-Task Learning without Demographic Information

Anonymous

TL;DR: We use a multi-task approach to transfer fairness from a task with demographic attributes to a new one without demographic attributes achieving fairness for the new task.

Abstract: Training supervised machine learning systems with a fairness loss can ensure prediction fairness across different demographic groups. However, doing so requires demographic annotations for training data, without which we cannot produce debiased classifiers for most tasks. Drawing inspiration from transfer learning methods, we investigate whether we can utilize demographic data from a related task to improve the fairness of a target task. We adapt a single-task fairness loss to a multi-task setting to exploit demographic labels from a related task in debiasing a target task, and demonstrate that demographic fairness objectives transfer fairness within a multi-task framework. Additionally, we show that this approach enables intersectional fairness by transferring between two datasets with different single-axis demographics. We explore different data domains to show how our loss can improve fairness domains and tasks.

Paper Type: long

Research Area: Ethics, Bias, and Fairness

Contribution Types: Model analysis & interpretability, NLP engineering experiment

Languages Studied: English