Nurse is Closer to Woman than Surgeon? Mitigating Gender-Biased Proximities in Word Embeddings

Nurse is Closer to Woman than Surgeon? Mitigating Gender-Biased Proximities in Word Embeddings Vaibhav Kumar1∗ Tenzin Singhay Bhotia1∗ Vaibhav Kumar1∗ Tanmoy Chakraborty2 1Delhi Technological University, New Delhi, India 2IIIT-Delhi, India 1{kumar.vaibhav1o1, tenzinbhotia0, vaibhavk992}@gmail.com 2tanmoy@iiitd.ac.in Abstract Word embeddings are the standard model for semanticandsyntactic representations of words. Unfortunately, these models have been shown to exhibit undesirable word associations result- ing from gender, racial, and religious biases. Existing post-processing methods for debias- ing word embeddings are unable to mitigate gender bias hidden in the spatial arrangement of word vectors. In this paper, we propose RAN-Debias, a novel gender debiasing meth- odology that not only eliminates the bias pre- sent in a word vector but also alters the spatial distribution of its neighboring vectors, achieving a bias-free setting while maintaining minimal semantic offset. We also propose a new bias evaluation metric, Gender-based Illicit Proximity Estimate (GIPE), which measures the extent of undue proximity in word vectors resulting from the presence of gender-based predilections. Experiments based on a suite of evaluation metrics show that RAN-Debias significantly outperforms the state-of-the-art in reducing proximity bias (GIPE) by at least 42.02%. It also reduces direct bias, adding minimal semantic distur- bance, and achieves the best performance in a downstream application task (coreference resolution). 1 Introduction Word embedding methods (Devlin et al., 2019; Mikolov et al., 2013a; Pennington et al., 2014) have been staggeringly successful in mapping the semantic space of words to a space of real-valued vectors, capturing both semantic and syntactic ∗Authors have contributed equally. relationships. However, as recent research has shown, word embeddings also possess a spectrum of biases related to gender (Bolukbasi et al., 2016; Hoyle et al., 2019), race, and religion (Manzini et al., 2019; Otterbacher et al., 2017). Bolukbasi et al. (2016) showed that there is a disparity in the association of professions with gender. For instance, while women are associated more closely with ‘‘receptionist’’ and ‘‘nurse’’, men are associated more closely with ‘‘doctor’’ and ‘‘engineer’’. Similarly, a word embedding model trained on data from a popular social media plat- form generates analogies such as ‘‘Muslim is to terrorist as Christian is to civilian’’ (Manzini et al., 2019). Therefore,given the large scale use of word embeddings, it becomes cardinal to remove the manifestation of biases. In this work, we focus on mitigating gender bias from pre-trained word embeddings. As shown in Table 1, the high degree of similarity between gender-biased words largely results from their individual proclivity towards a particular notion (gender in this case) rather than from empirical utility; we refer to such proximities as ‘‘illicit proximities’’. Existing debiasing methods (Bolukbasi et al., 2016; Kaneko and Bollegala, 2019) are primarily concerned with debiasing a word vector by mini- mising its projection on the gender direction. Although they successfully mitigate direct bias for a word, they tend to ignore the relationship bet- ween a gender-neutral word vector and its neighbors, thus failing to remove the gender bias encoded as illicit proximities between words (Gonen and Goldberg, 2019; Williams et al., 2019). For …

Content truncated to 500 words