Gender Bias in Coreference Resolution
TL
TL;DR
A novel, Winograd schema-style set of minimal pair sentences that differ only by pronoun gender are introduced, and systematic gender bias in three publicly-available coreference resolution systems is evaluated and confirmed.
Abstract
We present an empirical study of gender bias in coreference resolution systems. We first introduce a novel, Winograd schema-style set of minimal pair sentences that differ only by pronoun gender. With these “Winogender schemas,” we evaluate and confirm systematic gender bias in three publicly-available coreference resolution systems, and correlate this bias with real-world and textual gender statistics.
References27 items
1
Gender Bias in Coreference Resolution: Evaluation and Debiasing Methods
2
Mitigating Unwanted Biases with Adversarial Learning
3
Word embeddings quantify 100 years of gender and ethnic stereotypes
4
Men Also Like Shopping: Reducing Gender Bias Amplification using Corpus-level Constraints
5