Bolukbasi et al. (2016) presents one of the first gender bias mitigation techniques for word embeddings. Their method takes pre-trained word embeddings as input and attempts to isolate a linear subspace that captures most of the gender bias in the embeddings. As judged by an analogical evaluation task, their method virtually eliminates gender bias in the embeddings. However, an implicit and untested assumption of their method is that the bias sub-space is actually linear. In this work, we generalize their method to a kernelized, non-linear version. We take inspiration from kernel principal component analysis and derive a non-linear bias isolation technique. We discuss and overcome some of the practical drawbacks of our method for non-linear gender bias mitigation in word embeddings and analyze empirically whether the bias subspace is actually linear. Our analysis shows that gender bias is in fact well captured by a linear subspace, justifying the assumption of Bolukbasi et al. (2016).

中文翻译：

---

探索性别偏见缓解中的线性子空间假设

Bolukbasi 等。(2016) 提出了最早的词嵌入性别偏见缓解技术之一。他们的方法将预先训练的词嵌入作为输入，并尝试隔离一个线性子空间，该子空间捕获嵌入中的大部分性别偏见。根据类比评估任务判断，他们的方法实际上消除了嵌入中的性别偏见。然而，他们方法的一个隐含且未经测试的假设是偏置子空间实际上是线性的。在这项工作中，我们将他们的方法推广到内核化的非线性版本。我们从内核主成分分析中汲取灵感，推导出非线性偏置隔离技术。我们讨论并克服了我们的词嵌入中非线性性别偏见缓解方法的一些实际缺陷，并根据经验分析了偏见子空间是否实际上是线性的。我们的分析表明，性别偏见实际上被线性子空间很好地捕获，证明了 Bolukbasi 等人的假设是正确的。(2016)。