Mitigating Label Bias via Decoupled Confident Learning
This addresses algorithmic fairness issues for applications like content moderation by mitigating label bias, though it is incremental as it builds on existing bias mitigation methods.
The paper tackles the problem of label bias in training data, which is pervasive in domains like healthcare and hiring, by proposing Decoupled Confident Learning (DeCoLe) to identify and mitigate such bias, showing it outperforms competing approaches in hate speech detection.
Growing concerns regarding algorithmic fairness have led to a surge in methodologies to mitigate algorithmic bias. However, such methodologies largely assume that observed labels in training data are correct. This is problematic because bias in labels is pervasive across important domains, including healthcare, hiring, and content moderation. In particular, human-generated labels are prone to encoding societal biases. While the presence of labeling bias has been discussed conceptually, there is a lack of methodologies to address this problem. We propose a pruning method -- Decoupled Confident Learning (DeCoLe) -- specifically designed to mitigate label bias. After illustrating its performance on a synthetic dataset, we apply DeCoLe in the context of hate speech detection, where label bias has been recognized as an important challenge, and show that it successfully identifies biased labels and outperforms competing approaches.