Abstract | ||
---|---|---|
We introduce a new benchmark, WinoBias, for coreference resolution focused on gender bias. Our corpus contains Winograd-schema style sentences with entities corresponding to people referred by their occupation (e.g. the nurse, the doctor, the carpenter). We demonstrate that a rule-based, a feature-rich, and a neural coreference system all link gendered pronouns to pro-stereotypical entities with higher accuracy than anti-stereotypical entities, by an average difference of 21.1 in F1 score. Finally, we demonstrate a data-augmentation approach that, in combination with existing word-embedding debiasing techniques, removes the bias demonstrated by these systems in WinoBias without significantly affecting their performance on existing coreference benchmark datasets. Our dataset and code are available at this http URL |
Year | DOI | Venue |
---|---|---|
2018 | 10.18653/v1/N18-2003 | north american chapter of the association for computational linguistics |
DocType | Volume | Citations |
Journal | abs/1804.06876 | 10 |
PageRank | References | Authors |
0.60 | 13 | 5 |
Name | Order | Citations | PageRank |
---|---|---|---|
Jieyu Zhao | 1 | 50 | 5.89 |
Tianlu Wang | 2 | 38 | 4.68 |
Mark Yatskar | 3 | 176 | 11.14 |
Vicente Ordonez | 4 | 1418 | 69.65 |
Kai-Wei Chang | 5 | 4735 | 276.81 |