Evaluating Debiasing Techniques for Intersectional Biases
Evaluating Debiasing Techniques for Intersectional Biases
复制标题
评估交叉偏差的去偏技术
DOI：
--
复制
发表时间：
2021
期刊：
影响因子：
--
通讯作者：
Lea Frermann
中科院分区：
文献类型：
--
作者：
Shivashankar Subramanian;Xudong Han;Timothy Baldwin;Trevor Cohn;Lea Frermann
偏差在NLP模型中普遍存在，这推动了自动去偏技术的发展。对NLP去偏向方法的评估在很大程度上孤立地局限于二元属性，例如，关于二元性别或种族的去偏向，然而许多语料库涉及多个这样的属性，可能具有更高的基数。在这篇文章中，我们认为一个真正公平的模型必须考虑“重新划分选区”的群体，这些群体不仅包括单一属性，而且还包括交叉群体。我们评估了一种新的偏向约束模型，以及可以处理多个恒等式的迭代零空间投影技术的扩展。
Bias is pervasive for NLP models, motivating the development of automatic debiasing techniques. Evaluation of NLP debiasing methods has largely been limited to binary attributes in isolation, e.g., debiasing with respect to binary gender or race, however many corpora involve multiple such attributes, possibly with higher cardinality. In this paper we argue that a truly fair model must consider ‘gerrymandering’ groups which comprise not only single attributes, but also intersectional groups. We evaluate a form of bias-constrained model which is new to NLP, as well an extension of the iterative nullspace projection technique which can handle multiple identities.
黑人之于罪犯就像白人之于警察：检测和消除词嵌入中的多类偏差
DOI：
--
发表时间：
2019
期刊：
2019 Annual Conference of the North American Chapter of the Association for Computational Linguistics (NAACL
影响因子：
--
作者：
Manzini, Thomas;Lim, Yao Chong;Tsvetkov, Yulia;Black, Alan W
通讯作者：
Black, Alan W