A Robust Bias Mitigation Procedure Based on the Stereotype Content Model

Eddie Ungless, Amy Rafferty, Hrichika Nag, Björn Ross


Abstract
The Stereotype Content model (SCM) states that we tend to perceive minority groups as cold, incompetent or both. In this paper we adapt existing work to demonstrate that the Stereotype Content model holds for contextualised word embeddings, then use these results to evaluate a fine-tuning process designed to drive a language model away from stereotyped portrayals of minority groups. We find the SCM terms are better able to capture bias than demographic agnostic terms related to pleasantness. Further, we were able to reduce the presence of stereotypes in the model through a simple fine-tuning procedure that required minimal human and computer resources, without harming downstream performance. We present this work as a prototype of a debiasing procedure that aims to remove the need for a priori knowledge of the specifics of bias in the model.
Anthology ID:
2022.nlpcss-1.23
Volume:
Proceedings of the Fifth Workshop on Natural Language Processing and Computational Social Science (NLP+CSS)
Month:
November
Year:
2022
Address:
Abu Dhabi, UAE
Venue:
NLP+CSS
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
207–217
Language:
URL:
https://aclanthology.org/2022.nlpcss-1.23
DOI:
10.18653/v1/2022.nlpcss-1.23
Bibkey:
Cite (ACL):
Eddie Ungless, Amy Rafferty, Hrichika Nag, and Björn Ross. 2022. A Robust Bias Mitigation Procedure Based on the Stereotype Content Model. In Proceedings of the Fifth Workshop on Natural Language Processing and Computational Social Science (NLP+CSS), pages 207–217, Abu Dhabi, UAE. Association for Computational Linguistics.
Cite (Informal):
A Robust Bias Mitigation Procedure Based on the Stereotype Content Model (Ungless et al., NLP+CSS 2022)
Copy Citation:
PDF:
https://aclanthology.org/2022.nlpcss-1.23.pdf
Video:
 https://aclanthology.org/2022.nlpcss-1.23.mp4