Reducing Target Group Bias in Hate Speech Detectors
This addresses bias in hate speech detection for protected groups, which is an incremental improvement over existing methods.
The paper tackled the problem of bias in hate speech detectors, showing that models under-perform on protected groups like Black Women and Immigrants, with accuracy drops of 37% and 12% respectively. They proposed token-level hate sense disambiguation, reducing variance across target groups by at least 30% and improving worst-case performance by 13%.
The ubiquity of offensive and hateful content on online fora necessitates the need for automatic solutions that detect such content competently across target groups. In this paper we show that text classification models trained on large publicly available datasets despite having a high overall performance, may significantly under-perform on several protected groups. On the \citet{vidgen2020learning} dataset, we find the accuracy to be 37\% lower on an under annotated Black Women target group and 12\% lower on Immigrants, where hate speech involves a distinct style. To address this, we propose to perform token-level hate sense disambiguation, and utilize tokens' hate sense representations for detection, modeling more general signals. On two publicly available datasets, we observe that the variance in model accuracy across target groups drops by at least 30\%, improving the average target group performance by 4\% and worst case performance by 13\%.