What question did this study set out to answer?

The aim is to evaluate current explainable AI methods in hate speech detection to improve transparency and alignment with stakeholder needs.

Explainable AI for Hate Speech Moderation: A Stakeholder‐Centered and Sociotechnical Review

View Full Paper

Ask AI

Bookmark

Authors

This review synthesizes approaches for explainable AI in hate speech moderation, suggesting gaps and improvements for stakeholders.

The aim is to evaluate current explainable AI methods in hate speech detection to improve transparency and alignment with stakeholder needs.
Reviewed major explainable AI approaches, including ante-hoc, post hoc, local, global, counterfactual, and rationale-based methods.
Integrated perspectives from machine learning, human-computer interaction, and social sciences.
Proposed a conceptual framework mapping explanation strategies to stakeholder needs.
Identified gaps in dataset transparency, evaluation practices, and participatory design.
Highlighted significant gaps in the explainability of hate speech detection systems.
Emphasized the need for sociotechnical alignment and contextual sensitivity in AI systems.
Provided interdisciplinary recommendations for dataset development and inclusive design practices.

Qureshi et al. (2026) studied this question.

View Full Paper

Ask AI

Bookmark