iFair: Learning Individually Fair Data Representations for Algorithmic Decision Making

People are rated and ranked, towards algorithmic decision making in an increasing number of applications, typically based on machine learning. Research on how to incorporate fairness into such tasks has prevalently pursued the paradigm of group fairness: ensuring that each ethnic or social group receives its fair share in the outcome of classifiers and rankings. In contrast, the alternative paradigm of individual fairness has received relatively little attention. This paper introduces a method for probabilistically clustering user records into a low-rank representation that captures individual fairness yet also achieves high accuracy in classification and regression models. Our notion of individual fairness requires that users who are similar in all task-relevant attributes such as job qualification, and disregarding all potentially discriminating attributes such as gender, should have similar outcomes. Since the case for fairness is ubiquitous across many tasks, we aim to learn general representations that can be applied to arbitrary downstream use-cases. We demonstrate the versatility of our method by applying it to classification and learning-to-rank tasks on two real-world datasets. Our experiments show substantial improvements over the best prior work for this setting.