The effects of gender bias in word embeddings on depression prediction

TL;DR

This paper investigates gender bias in word embeddings used for depression prediction, showing how biases vary across embedding types and can be mitigated through gender word swapping.

Contribution

It provides a detailed analysis of gender bias in various pre-trained embeddings within the mental health domain and proposes a simple data augmentation method to reduce bias.

Findings

Bias varies with embedding type and training data.

Gender swapping reduces bias in depression prediction.

Bias transfer affects downstream mental health tasks.

Abstract

Word embeddings are extensively used in various NLP problems as a state-of-the-art semantic feature vector representation. Despite their success on various tasks and domains, they might exhibit an undesired bias for stereotypical categories due to statistical and societal biases that exist in the dataset they are trained on. In this study, we analyze the gender bias in four different pre-trained word embeddings specifically for the depression category in the mental disorder domain. We use contextual and non-contextual embeddings that are trained on domain-independent as well as clinical domain-specific data. We observe that embeddings carry bias for depression towards different gender groups depending on the type of embeddings. Moreover, we demonstrate that these undesired correlations are transferred to the downstream task for depression phenotype recognition. We find that data augmentation by simply swapping gender words mitigates the bias significantly in the downstream task.