arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 1809.01496cs.CLcs.LGstat.ML
收藏

学习中性词嵌入

Learning Gender-Neutral Word Embeddings

University of California, Los Angeles(加州大学洛杉矶分校)

Jieyu Zhao, Yichao Zhou, Zeyu Li, Wei Wang, Kai-Wei Chang

摘要:

词嵌入模型已经成为广泛的自然语言处理(NLP)应用中的基本组件。然而,在人类生成的语料库上训练的嵌入已被证明继承了反映社会结构的强烈性别刻板印象。为了解决这个问题,在本文中,我们提出了一个新的训练过程学习性别中立的词嵌入。我们的方法旨在保留词向量的某些维度中的性别信息,同时迫使其他维度不受性别影响。基于所提出的方法,我们生成GloVe的性别中立的变体(GN-GloVe)。定量和定性实验表明,GN-GloVe成功地隔离了性别信息,而不牺牲嵌入模型的功能。

英文摘要:

Word embedding models have become a fundamental component in a wide range of Natural Language Processing (NLP) applications. However, embeddings trained on human-generated corpora have been demonstrated to inherit strong gender stereotypes that reflect social constructs. To address this concern, in this paper, we propose a novel training procedure for learning gender-neutral word embeddings. Our approach aims to preserve gender information in certain dimensions of word vectors while compelling other dimensions to be free of gender influence. Based on the proposed method, we generate a Gender-Neutral variant of GloVe (GN-GloVe). Quantitative and qualitative experiments demonstrate that GN-GloVe successfully isolates gender information without sacrificing the functionality of the embedding model.

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。

↑