A machine learning project to detect Indonesian-language YouTube comments related to online gambling using a IndoBERT model.

How to Run

1. Clone the Repository

git clone https://github.com/HAJAR-Enterprise/ML-Repo.git
cd ML-Repo

2. Install Requirements

pip install -r requirements.txt

3. Run the Notebooks

Open and run the notebooks in the following order:

• youtube_comment_scraper.ipynb → to scrape YouTube comments
• data_preprocessing.ipynb → to clean and preprocess data
• modelling.ipynb → to train and evaluate the model

Evaluation Summary

• Model Accuracy: 99%
• F1-Score: 0.99
• Dataset Size: 15,000 Indonesian comments
• Can detect disguised or symbol-altered gambling terms automatically

Note

youtube_comment_scraper.ipynb included API credentials.