Towards Understanding Gender Bias in Relation Extraction
Abstract
Recent developments in Neural Relation Extraction (NRE) have made significant strides towards Automated Knowledge Base Construction (AKBC). While much attention has been dedicated towards improvements in accuracy, there have been no attempts in the literature to our knowledge to evaluate social biases in NRE systems. We create WikiGenderBias, a distantly supervised dataset with a human annotated test set. WikiGenderBias has sentences specifically curated to analyze gender bias in relation extraction systems. We use WikiGenderBias to evaluate systems for bias and find that NRE systems exhibit gender biased predictions and lay groundwork for future evaluation of bias in NRE. We also analyze how name anonymization, hard debiasing for word embeddings, and counterfactual data augmentation affect gender bias in predictions and performance.
Cite
@article{arxiv.1911.03642,
title = {Towards Understanding Gender Bias in Relation Extraction},
author = {Andrew Gaut and Tony Sun and Shirlyn Tang and Yuxin Huang and Jing Qian and Mai ElSherief and Jieyu Zhao and Diba Mirza and Elizabeth Belding and Kai-Wei Chang and William Yang Wang},
journal= {arXiv preprint arXiv:1911.03642},
year = {2020}
}