# Debiasing Embeddings for Reduced Gender Bias in Text Classification

## TL;DR

This paper examines how debiasing word embeddings impacts gender bias in text classification, revealing that traditional methods may worsen bias but can be adjusted to reduce bias while preserving accuracy.

## Contribution

It identifies limitations of existing debiasing techniques and proposes a simple adjustment to effectively reduce gender bias without sacrificing classification performance.

## Key findings

- Traditional debiasing can increase downstream bias
- Adjusted techniques reduce bias and maintain accuracy
- Debiasing impacts are task-dependent

## Abstract

(Bolukbasi et al., 2016) demonstrated that pretrained word embeddings can inherit gender bias from the data they were trained on. We investigate how this bias affects downstream classification tasks, using the case study of occupation classification (De-Arteaga et al.,2019). We show that traditional techniques for debiasing embeddings can actually worsen the bias of the downstream classifier by providing a less noisy channel for communicating gender information. With a relatively minor adjustment, however, we show how these same techniques can be used to simultaneously reduce bias and maintain high classification accuracy.