IndoBERT emotion classification
Classifying emotion in Indonesian text with an IndoBERT model.
- Focus
- Natural language processing
- Role / project
- Data science · XPLORE
- Tools / approach
- Python, IndoBERT, PyTorch, Hugging Face Transformers
Context
This model was developed as part of XPLORE, a Pustakadata project for monitoring social media and news. It focuses on classifying emotion in Indonesian-language text.
The work
I developed the emotion classifier as one of four machine learning components in XPLORE. The published model uses BERT sequence classification with six labels: sadness, anger, love, fear, happy, and neutral. It can be loaded through the Hugging Face Transformers text-classification pipeline.
Outcome
A publicly available model with configuration, model weights, and Python usage examples on Hugging Face. Other developers can load it into their own applications or notebooks using Transformers.