Hidden or Inferred: Fair Learning-To-Rank with Unknown Demographics

TL;DR

This paper investigates how errors in inferring sensitive demographic attributes affect the fairness of learning-to-rank models, comparing strategies that use hidden, inferred, or no demographic data, and finds re-ranking methods are more robust to inference errors.

Contribution

It provides a systematic empirical analysis of the impact of demographic inference errors on fair learning-to-rank strategies, highlighting the robustness of re-ranking approaches.

Findings

Fairness-aware LTR methods can increase bias with inference noise.

Fair re-ranking strategies are more robust to demographic inference errors.

Inference errors significantly impact the fairness performance of LTR models.

Abstract

As learning-to-rank models are increasingly deployed for decision-making in areas with profound life implications, the FairML community has been developing fair learning-to-rank (LTR) models. These models rely on the availability of sensitive demographic features such as race or sex. However, in practice, regulatory obstacles and privacy concerns protect this data from collection and use. As a result, practitioners may either need to promote fairness despite the absence of these features or turn to demographic inference tools to attempt to infer them. Given that these tools are fallible, this paper aims to further understand how errors in demographic inference impact the fairness performance of popular fair LTR strategies. In which cases would it be better to keep such demographic attributes hidden from models versus infer them? We examine a spectrum of fair LTR strategies ranging from fair LTR with and without demographic features hidden versus inferred to fairness-unaware LTR followed by fair re-ranking. We conduct a controlled empirical investigation modeling different levels of inference errors by systematically perturbing the inferred sensitive attribute. We also perform three case studies with real-world datasets and popular open-source inference methods. Our findings reveal that as inference noise grows, LTR-based methods that incorporate fairness considerations into the learning process may increase bias. In contrast, fair re-ranking strategies are more robust to inference errors. All source code, data, and experimental artifacts of our experimental study are available here: https://github.com/sewen007/hoiltr.git