Hate Speech Classification And Categorization

via “hate-speech-and-discrimination-detection”

Google's safety content classifiers built on Gemma.

Unique: Provides multi-dimensional categorization (hate speech type + target group) rather than binary classification, enabling granular moderation policies. Gemma's semantic understanding captures coded language and dog whistles beyond simple keyword matching.

vs others: More nuanced than regex-based slur filters because it understands context and coded language; more deployable than cloud APIs because it runs on-device with no external dependencies

Unique: Uses keyword-to-category mapping with pattern rules to classify hate speech into discrete categories, enabling policy-driven moderation workflows. This is more operationally transparent than black-box ML models but less adaptable to emerging hate speech patterns.

vs others: More transparent and auditable than ML-based classifiers for compliance purposes, but less accurate at detecting novel or subtle hate speech compared to fine-tuned transformer models like those in Perspective API.

via “multilingual hate speech classification”