Abstract
Gender bias exists in natural language datasets, which neural language models tend to learn, resulting in biased text generation. In this research, we propose a debiasing approach based on the loss function modification. We introduce a new term to the loss function which attempts to equalize the probabilities of male and female words in the output. Using an array of bias evaluation metrics, we provide empirical evidence that our approach successfully mitigates gender bias in language models without increasing perplexity. In comparison to existing debiasing strategies, data augmentation, and word embedding debiasing, our method performs better in several aspects, especially in reducing gender bias in occupation words. Finally, we introduce a combination of data augmentation and our approach and show that it outperforms existing strategies in all bias evaluation metrics.- Anthology ID:
- P19-2031
- Volume:
- Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics: Student Research Workshop
- Month:
- July
- Year:
- 2019
- Address:
- Florence, Italy
- Editors:
- Fernando Alva-Manchego, Eunsol Choi, Daniel Khashabi
- Venue:
- ACL
- SIG:
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 223–228
- Language:
- URL:
- https://aclanthology.org/P19-2031
- DOI:
- 10.18653/v1/P19-2031
- Cite (ACL):
- Yusu Qian, Urwa Muaz, Ben Zhang, and Jae Won Hyun. 2019. Reducing Gender Bias in Word-Level Language Models with a Gender-Equalizing Loss Function. In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics: Student Research Workshop, pages 223–228, Florence, Italy. Association for Computational Linguistics.
- Cite (Informal):
- Reducing Gender Bias in Word-Level Language Models with a Gender-Equalizing Loss Function (Qian et al., ACL 2019)
- PDF:
- https://preview.aclanthology.org/nschneid-patch-3/P19-2031.pdf
- Code
- additional community code