Man is to computer programmer as woman is to homemaker? Debiasing word embeddings

Paper Notes

Bolukbasi, Tolga, et al. “Man is to computer programmer as woman is to homemaker? Debiasing word embeddings.” Advances in Neural Information Processing Systems. 2016.

[pdf]


Contributions

  • Eliminate gender biases in word embeddings
    • Remove Receptionist-Female; gender neutral words
    • Keep Queen-Female; generder specific words
  • Prove linear separable between gender-neutral words and baised words
  • Algorithm to “debias”

Examples

man - woman = compupter_programmer - homemaker

Details

  • wordvec on Google News, 300d
  • GloVe is tested and described in Appendix

Methods

  • Gender occupation stereotype exists!
    • M-Turk
  • Analogies
    • (she, he) -> (x, y)`
    • Overall, 72 out of 150 analogies were rated as gender-appropriate by five or more out of 10 crowd-workers, and 29 analogies were rated as exhibiting gender stereotype by five or more crowd-workers
    • The first principal component capture the gender subspace.

      Image of Yaktocat

    • Gender direction? is it the corresponding eigenvector?
    • DirectBiase=0.8 -> high?
  • Debiasing Algorithm
  • identify gender subspace
  • hard debiasing / soft debiasing
  • Human evaluation

  1. Zhao, Jieyu, et al. “Men Also Like Shopping: Reducing Gender Bias Amplification using Corpus-level Constraints.” arXiv preprint arXiv:1707.09457 (2017).
Written on October 4, 2017