New dataset ChildGuard targets child-specific hate speech on social media

Researchers have introduced ChildGuard, a new dataset designed to combat child-targeted hate speech on social media platforms. This dataset contains over 350,000 annotated instances collected from X, Reddit, and YouTube, categorized by age groups including younger children, pre-teens, and teens. Evaluations using transformer-based models and LLMs on ChildGuard achieved a Macro-F1 score of 82.07%, highlighting the dataset's utility in developing more effective detection systems for age-specific hate speech. AI

IMPACT Enhances AI's ability to detect and mitigate child-specific online hate speech, improving platform safety.

RANK_REASON Research paper introducing a new dataset for a specific NLP task. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CL →

• arXiv
• ChildGuard
• Gautam Siddharth Kashyap
• Hugging Face
• Reddit
• X
• YouTube

• paper
• safety

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →