SPAR-Hate: An Auditor-Guided Multi-Agent Framework for Bilingual Hate Speech Parsing

Abstract:Hate speech detection has recently shifted from coarse-grained classification to structured parsing, where systems must jointly identify hateful targets, arguments, and target-level labels. However, existing studies primarily emphasize benchmark evaluation while paying less attention to the cultural, linguistic, and social-group challenges involved in structured hate speech parsing. To address these challenges, we propose SPAR-Hate, an auditor-guided multi-agent framework for bilingual hate speech parsing. The framework first decomposes documents into clause-level decision units and then generates evidence-grounded judgments from three complementary perspectives: Victim, Moderator, and Cultural Bystander. An evidence-constrained arbitration process resolves conflicts among role-specific predictions and aggregates them into structured sample-level outputs. Experiments on the STATE-ToxiCN and TBO benchmarks show that SPAR-Hate consistently improves bilingual hate parsing across diverse large language models. The framework achieves state-of-the-art results on bilingual multi-tuple extraction tasks, with the largest gains observed under stricter structural evaluation metrics.

16 pages, 2 figures. Submitted to EMNLP 2026

Artificial Intelligence (cs.AI)

arXiv:2608.22018 [cs.AI]

(or arXiv:2608.22018v1 [cs.AI] for this version)

https://doi.org/10.48550/arXiv.2608.22018 Focus to learn more arXiv-issued DOI via DataCite (pending registration)

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.