论文标题
MDR群集 - debias：非线性文字贴式偏见管道
MDR Cluster-Debias: A Nonlinear WordEmbedding Debiasing Pipeline
论文作者
论文摘要
现有的词汇嵌入方法通常仅在表面上才能做到这一点，而与与原始嵌入空间中的特定性别相关的单词，仍然可以将其聚集在词汇空间中。但是，尚未有一项研究探讨为什么存在这种残留的聚类以及如何解决。目前的工作填补了这一空白。我们确定存在残留偏差的两个潜在原因，并开发新的管道MDR群集 - debias来减轻这种偏见。我们探索了我们方法的优势和劣势，发现它在各种上游偏见测试上大大优于其他现有的偏见方法，但是在下游任务中减少性别偏见方面的改善有限。这表明单词嵌入方式以其他方式编码性别偏差，不一定会被上游测试捕获。
Existing methods for debiasing word embeddings often do so only superficially, in that words that are stereotypically associated with, e.g., a particular gender in the original embedding space can still be clustered together in the debiased space. However, there has yet to be a study that explores why this residual clustering exists, and how it might be addressed. The present work fills this gap. We identify two potential reasons for which residual bias exists and develop a new pipeline, MDR Cluster-Debias, to mitigate this bias. We explore the strengths and weaknesses of our method, finding that it significantly outperforms other existing debiasing approaches on a variety of upstream bias tests but achieves limited improvement on decreasing gender bias in a downstream task. This indicates that word embeddings encode gender bias in still other ways, not necessarily captured by upstream tests.
