Our work focuses on the biases that emerge in the natural language generation (NLG) task of sentence completion. In this paper, we introduce a framework of fairness for NLG followed by an evaluation of gender biases in two state-of-the-art language models. Our analysis provides a theoretical formulation for biases in NLG and empirical evidence that existing language generation models embed gender bias.
@article{arxiv.2008.01548,
title = {Defining and Evaluating Fair Natural Language Generation},
author = {Catherine Yeo and Alyssa Chen},
journal= {arXiv preprint arXiv:2008.01548},
year = {2020}
}
Comments
7 pages, 2 figures, to be published in Proceedings of the The Fourth Widening Natural Language Processing Workshop at ACL