Debiasing Gender Bias In Information Retrieval Models

Abstract

Biases in culture, gender, ethnicity, etc. have existed for decades and have affected many areas of human social interaction. These biases have been shown to impact machine learning (ML) models, and for natural language processing (NLP), this can have severe consequences for downstream tasks. Mitigating gender bias in information retrieval (IR) is important to avoid propagating stereotypes. In this work, we employ a dataset consisting of two components: (1) relevance of a document to a query and (2) "gender" of a document, in which pronouns are replaced by male, female, and neutral conjugations. We definitively show that pre-trained models for IR do not perform well in zero-shot retrieval tasks when full fine-tuning of a large pre-trained BERT encoder is performed and that lightweight fine-tuning performed with adapter networks improves zero-shot retrieval performance almost by 20% over baseline. We also illustrate that pre-trained models have gender biases that result in retrieved art

Stats

• citations0
• S2 citations—
• github stars0
• HF likes0
• heat score0.00
• arxiv keysundararaman2022debiasing

Related papers

• Do Neural Ranking Models Intensify Gender Bias? (2020)12.47
• Mitigating Test-time Bias For Fair Image Retrieval (2023)0.00
• Writing Style Matters: An Examination Of Bias And Fairness In Information Retrieval Systems (2024)4.52
• An Empirical Study Of Position Bias In Modern Information Retrieval (2025)1.69
• Hard Negatives, Hard Lessons: Revisiting Training Data Quality For Robust Information Retrieval With Llms (2025)2.26
• BEIR: A Heterogenous Benchmark For Zero-shot Evaluation Of Information Retrieval Models (2021)6.67
• Posir: Position-aware Heterogeneous Information Retrieval Benchmark (2026)0.00
• Invisible Relevance Bias: Text-image Retrieval Models Prefer Ai-generated Images (2023)9.23