arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2010.02458cs.LGcs.CLcs.IR
收藏

鲁棒文本分类中虚假相关的识别

Identifying Spurious Correlations for Robust Text Classification

Illinois Institute of Technology(伊利诺伊理工学院);Tulane University(杜兰大学)

Zhao Wang, Aron Culotta

摘要:

文本分类器的预测通常是由虚假的相关性驱动的,例如,术语“斯皮尔伯格”与正面评论的电影相关,即使该术语本身在语义上并不传达正面的情感。在本文中,我们提出了一种方法来区分虚假和真正的相关性在文本分类。我们将其视为一个监督分类问题,使用来自治疗效果估计的特征来区分虚假的相关性与“真正的”相关性。由于这些功能的通用性和它们的小维度,我们发现,即使有限的训练样本的方法工作得很好,它是可能的运输词分类到新的领域。在四个数据集(情感分类和毒性检测)上的实验表明,使用这种方法来通知特征选择也会导致更鲁棒的分类,这是通过改善受虚假相关影响的样本的最坏情况准确性来衡量的。

英文摘要:

The predictions of text classifiers are often driven by spurious correlations -- e.g., the term `Spielberg' correlates with positively reviewed movies, even though the term itself does not semantically convey a positive sentiment. In this paper, we propose a method to distinguish spurious and genuine correlations in text classification. We treat this as a supervised classification problem, using features derived from treatment effect estimators to distinguish spurious correlations from "genuine" ones. Due to the generic nature of these features and their small dimensionality, we find that the approach works well even with limited training examples, and that it is possible to transport the word classifier to new domains. Experiments on four datasets (sentiment classification and toxicity detection) suggest that using this approach to inform feature selection also leads to more robust classification, as measured by improved worst-case accuracy on the samples affected by spurious correlations.

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。

↑