Academic Project
Banglish Cyberbullying Detection
On-device ML model that detects cyberbullying in Banglish (Bengali written in Latin script) — runs at the edge without sending data to a server.
PythonEdge AI / MLNLP
The problem
Bengali-speaking communities use Banglish on social platforms, and existing English-trained moderation models miss it entirely. Sending posts to a cloud classifier raises privacy and latency concerns.
Architecture
- Custom-trained NLP model on Banglish text corpora
- Edge inference for on-device classification — no server round-trip
Highlights
Banglish-native NLP
Trained on Bengali-written-in-Latin-script corpora that English-trained moderation models miss entirely — built for how Bengali communities actually post.
On-device edge inference
Classification runs locally with no server round-trip, so posts never leave the device — better privacy and lower latency than a cloud classifier.