The problem
People write by mixing languages and scripts, and most language tools fail on this natural style.
What to build
Build a model or pipeline that understands code-mixed messages for tasks such as intent detection and sentiment.
Success looks like
Competitive scores on a public code-mixed benchmark with an error analysis by language pair.
Suggested technologies
- Transformers
- Code-mixed datasets
- Transliteration
- Evaluation
- Fine-tuning
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.