The Undesirable Dependence on Frequency of Gender Bias Metrics Based on Word Embeddings

Francisco Valentini, Germán Rosati, Diego Fernandez Slezak, Edgar Altszyler


Abstract
Numerous works use word embedding-based metrics to quantify societal biases and stereotypes in texts. Recent studies have found that word embeddings can capture semantic similarity but may be affected by word frequency. In this work we study the effect of frequency when measuring female vs. male gender bias with word embedding-based bias quantification methods. We find that Skip-gram with negative sampling and GloVe tend to detect male bias in high frequency words, while GloVe tends to return female bias in low frequency words. We show these behaviors still exist when words are randomly shuffled. This proves that the frequency-based effect observed in unshuffled corpora stems from properties of the metric rather than from word associations. The effect is spurious and problematic since bias metrics should depend exclusively on word co-occurrences and not individual word frequencies. Finally, we compare these results with the ones obtained with an alternative metric based on Pointwise Mutual Information. We find that this metric does not show a clear dependence on frequency, even though it is slightly skewed towards male bias across all frequencies.
Anthology ID:
2022.findings-emnlp.373
Volume:
Findings of the Association for Computational Linguistics: EMNLP 2022
Month:
December
Year:
2022
Address:
Abu Dhabi, United Arab Emirates
Editors:
Yoav Goldberg, Zornitsa Kozareva, Yue Zhang
Venue:
Findings
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
5086–5092
Language:
URL:
https://aclanthology.org/2022.findings-emnlp.373
DOI:
10.18653/v1/2022.findings-emnlp.373
Bibkey:
Cite (ACL):
Francisco Valentini, Germán Rosati, Diego Fernandez Slezak, and Edgar Altszyler. 2022. The Undesirable Dependence on Frequency of Gender Bias Metrics Based on Word Embeddings. In Findings of the Association for Computational Linguistics: EMNLP 2022, pages 5086–5092, Abu Dhabi, United Arab Emirates. Association for Computational Linguistics.
Cite (Informal):
The Undesirable Dependence on Frequency of Gender Bias Metrics Based on Word Embeddings (Valentini et al., Findings 2022)
Copy Citation:
PDF:
https://aclanthology.org/2022.findings-emnlp.373.pdf
Software:
 2022.findings-emnlp.373.software.zip