BriefGPT.xyz
May, 2019
匿名 BERT：性别代词解析挑战的一种增强方法
Anonymized BERT: An Augmentation Approach to the Gendered Pronoun Resolution Challenge
HTML
PDF
Bo Liu
TL;DR
使用BERT模型和新颖的数据增强策略，利用共同的填充名称对候选名字进行匿名化处理，提高数据量的同时使模型更好地识别名称、缩短标记长度，并消除与名称相关的性别和地域偏见，有效提升了Gendered Pronoun Resolution挑战的性能。
Abstract
We present our 7th place solution to the
gendered pronoun resolution
challenge, which uses
bert
without fine-tuning and a novel
augmentation stra
→