Nationality Bias in Text Generation

作者：Pranav Narayanan Venkit, Sanjana Gautam, Ruchi Panchanadikar, Ting-Hao 'Kenneth' Huang, Shomir Wilson · 发表于：Conference of the European Chapter of the Association for Computational Linguistics · 年份：2023 · DOI：10.48550/arxiv.2302.02463 · 被引用次数：84 · 研究领域：Computer Science

Little attention is placed on analyzing nationality bias in language models, especially when nationality is highly used as a factor in increasing the performance of social NLP models. This paper examines how a text generation model, GPT-2, accentuates pre-existing societal biases about country-based demonyms. We generate stories using GPT-2 for various nationalities and use sensitivity analysis to explore how the number of internet users and the country’s economic status impacts the sentiment of the stories. To reduce the propagation of biases through large language models (LLM), we explore the debiasing method of adversarial triggering. Our results show that GPT-2 demonstrates significant bias against countries with lower internet users, and adversarial triggering effectively reduces the same.