Corpus ID: 1704893

Man is to Computer Programmer as Woman is to Homemaker? Debiasing Word Embeddings

  title={Man is to Computer Programmer as Woman is to Homemaker? Debiasing Word Embeddings},
  author={Tolga Bolukbasi and Kai-Wei Chang and James Y. Zou and Venkatesh Saligrama and A. Kalai},
  • Tolga Bolukbasi, Kai-Wei Chang, +2 authors A. Kalai
  • Published 2016
  • Computer Science, Mathematics
  • ArXiv
  • The blind application of machine learning runs the risk of amplifying biases present in data. [...] Key Method Geometrically, gender bias is first shown to be captured by a direction in the word embedding. Second, gender neutral words are shown to be linearly separable from gender definition words in the word embedding.Expand Abstract
    884 Citations
    Fair is Better than Sensational: Man is to Doctor as Woman is to Doctor
    • 27
    • Highly Influenced
    • PDF
    Lipstick on a Pig: Debiasing Methods Cover up Systematic Gender Biases in Word Embeddings But do not Remove Them
    • 145
    • Highly Influenced
    • PDF
    Understanding the Origins of Bias in Word Embeddings
    • 35
    • Highly Influenced
    • PDF
    Equalizing Gender Biases in Neural Machine Translation with Word Embeddings Techniques
    • 40
    • Highly Influenced
    • PDF
    Nurse is Closer to Woman than Surgeon? Mitigating Gender-Biased Proximities in Word Embeddings
    • 1
    • Highly Influenced
    • PDF
    Bias in word embeddings
    • 10
    • Highly Influenced
    "Thy algorithm shalt not bear false witness": An Evaluation of Multiclass Debiasing Methods on Word Embeddings
    Humpty Dumpty: Controlling Word Meanings via Corpus Poisoning
    • 7
    • PDF


    Glove: Global Vectors for Word Representation
    • 15,494
    • PDF
    Linguistic Regularities in Continuous Space Word Representations
    • 2,568
    • PDF
    Linguistic Regularities in Sparse and Explicit Word Representations
    • 532
    • PDF
    Distributed Representations of Words and Phrases and their Compositionality
    • 19,642
    • PDF
    Deep Recursive Neural Networks for Compositionality in Language
    • 225
    • PDF
    Improving Document Ranking with Dual Word Embeddings
    • 120
    • PDF
    Automated Experiments on Ad Privacy Settings
    • 309
    • PDF
    Certifying and Removing Disparate Impact
    • 678
    • PDF
    Semi-supervised Question Retrieval with Gated Convolutions
    • 66
    • PDF
    It's a Man's Wikipedia? Assessing Gender Inequality in an Online Encyclopedia
    • 114
    • PDF