What does this research mean for the field?

An enhanced two-stage learning model incorporating multi-class rationale learning and multi-task learning improves the explainability of hate speech detection while maintaining competitive detection performance. Novelty: ClaimNovelty.METHODOLOGICAL. Consensus alignment: ConsensusAlignment.NEUTRAL.

What question did this study set out to answer?

Develop an explainable hate speech detection framework that generates clear semantic rationales without compromising classification accuracy.

August 23, 2026International Journal of Web Information Systems

Explainable hate speech detection with multi-class rationale learning and multi-task learning

View Full Paper

Ask AI

Bookmark

Authors

雄本雄太本荘Iwate University JZJianwei ZhangIwate University WZWei ZhangHuazhong University of Science and Technology

Discussion

Member takes

Implication

Computational study demonstrates improved explainability and robust accuracy in automated hate speech detection, indicating the value of joint rationale and target modeling.

Key Points

Develop an explainable hate speech detection framework that generates clear semantic rationales without compromising classification accuracy.
Designed a two-stage architecture featuring rationale learning in the initial stage followed by model fine-tuning.
Incorporated multi-class rationale learning to extract fine-grained semantics alongside multi-task learning for joint hate speech detection and target classification.
Evaluated explainability quality and classification effectiveness on English-language hate speech benchmarks.
The proposed two-stage framework enhanced prediction explainability while maintaining competitive hate speech detection performance.
Ablation analyses demonstrated that both multi-class rationale extraction and target-oriented multi-task learning independently improved explanation quality.

Cite This Study

本荘 et al. (2026) studied this question.

synapsesocial.com/papers/6a8aae5c7677a3411444752e https://doi.org/10.1108/ijwis-03-2026-0154

View Full Paper

Ask AI

Bookmark

Implication

Computational study demonstrates improved explainability and robust accuracy in automated hate speech detection, indicating the value of joint rationale and target modeling.

Explainable hate speech detection with multi-class rationale learning and multi-task learning

Discussion

Member takes

Implication

Key Points

Cite This Study

Explainable hate speech detection with multi-class rationale learning and multi-task learning

Discussion

Member takes

Implication

Key Points

Cite This Study

Also Consider

Also Consider