arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1904.03310cs.CL
收藏

语境化词嵌入中的性别偏见

Gender Bias in Contextualized Word Embeddings

Allen Institute for Artificial Intelligence(艾伦人工智能研究所);University of California, Los Angeles(加州大学洛杉矶分校);University of Virginia(弗吉尼亚大学);University of Cambridge(剑桥大学)

Jieyu Zhao, Tianlu Wang, Mark Yatskar, Ryan Cotterell, Vicente Ordonez, Kai-Wei Chang

摘要:

在本文中,我们量化,分析和减轻ELMo的语境化词向量中表现出的性别偏见。首先,我们进行了几个内在的分析,发现(1)训练数据ELMo包含显着更多的男性比女性实体,(2)训练ELMo嵌入系统编码性别信息和(3)ELMo不平等地编码性别信息的男性和女性实体。然后,我们表明,一个国家的最先进的共指系统,依赖于ELMo继承其偏见,并表现出显着的偏见WinoBias探测语料库。最后,我们探索了两种方法来减轻这种性别偏见,并表明WinoBias上的偏见可以消除。

英文摘要:

In this paper, we quantify, analyze and mitigate gender bias exhibited in ELMo's contextualized word vectors. First, we conduct several intrinsic analyses and find that (1) training data for ELMo contains significantly more male than female entities, (2) the trained ELMo embeddings systematically encode gender information and (3) ELMo unequally encodes gender information about male and female entities. Then, we show that a state-of-the-art coreference system that depends on ELMo inherits its bias and demonstrates significant bias on the WinoBias probing corpus. Finally, we explore two methods to mitigate such gender bias and show that the bias demonstrated on WinoBias can be eliminated.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑