Necessity and Sufficiency for Explaining Text Classifiers: A Case Study in Hate Speech Detection. (arXiv:2205.03302v1 [cs.CL])

Web: http://arxiv.org/abs/2205.03302

We present a novel feature attribution method for explaining text
classifiers, and analyze it in the context of hate speech detection. Although
feature attribution models usually provide a single importance score for each
token, we instead provide two complementary and theoretically-grounded scores
-- necessity and sufficiency -- resulting in more informative explanations. We
propose a transparent method that calculates these values by generating
explicit perturbations of the input text, allowing the importance scores
themselves to be explainable. We employ our …

arxiv case study detection hate speech speech study text