metadata
language:
- ms
tags:
- cyberbullying
- abusive-language
- malay
- classification
license: mit
datasets:
- mohanrj/MYBully
metrics:
- accuracy
- f1
base_model:
- mesolitica/roberta-base-bahasa-cased
MYBully-CyBERT (Manual)
Model Overview
This model is MYBully-CyBERT, fine-tuned exclusively on the manually annotated portion of the MYBully dataset.
It is designed for Cyberbullying Detection in Bahasa Malaysia tweets.
Intended Use
- Cyberbullying and abusive language detection (binary classification).
Training Data
- Dataset: MYBully (Bahasa Malaysia tweets).
- Annotation: Manual
Model Details
- Base model: roberta-base-bahasa-cased
- Fine-tuning: Binary classification head
- Labels: Cyberbully (1) vs Non-cyberbully (0)
Performance
| Metric | Value |
|---|---|
| Accuracy | 0.84 |
| Precision | 0.85 |
| Recall | 0.85 |
| F1 | 0.85 |