The problem
Moderation tools trained on English miss abuse in regional languages and slang.
What to build
Develop a classifier and review workflow for abusive content in at least two Indian languages, with a bias analysis.
Success looks like
Reported precision and recall per language, a documented bias study and an appeals flow.
Suggested technologies
- Text classification
- Low-resource NLP
- Bias evaluation
- Annotation
- Human review
An original, independent statement. It is not an official Smart India Hackathon problem statement, and Arventiq Labs is not affiliated with or endorsed by that programme. The scenario is illustrative and does not describe any specific organisation. Free to use with credit.