A multi-label toxicity dataset for Nepali social media content.
Contains 555 samples covering hate speech, casteism, religious
incitement, political defamation, and cyberbullying in Nepali,
Romanized Nepali, Mixed, and English text.
hate_speech: Attacks on groups based on identity
casteism: Caste based discrimination
religious_incitement: Attacks on religion or religious groups
political_defamation: False malicious attacks on… See the full description on the dataset page:
https://huggingface.co/datasets/biraj-bhusal/rakshak-nepali-toxicity.
