Resolving Gendered Ambiguous Pronouns with BERT
This addresses a fairness issue in coreference resolution for NLP applications like machine translation and chatbots, though it is incremental as it builds on BERT.
The paper tackled the problem of gender-biased pronoun resolution in NLP, achieving a 92% F1 score and reduced gender bias on a benchmark dataset.
Pronoun resolution is part of coreference resolution, the task of pairing an expression to its referring entity. This is an important task for natural language understanding and a necessary component of machine translation systems, chat bots and assistants. Neural machine learning systems perform far from ideally in this task, reaching as low as 73% F1 scores on modern benchmark datasets. Moreover, they tend to perform better for masculine pronouns than for feminine ones. Thus, the problem is both challenging and important for NLP researchers and practitioners. In this project, we describe our BERT-based approach to solving the problem of gender-balanced pronoun resolution. We are able to reach 92% F1 score and a much lower gender bias on the benchmark dataset shared by Google AI Language team.