Detecting and Preventing Bad and Vietnamese Hate and Offensive Content

DePrBViHO is a tool designed to identify and prevent bad and offensive content in the Vietnamese language. The project utilizes advanced natural language processing techniques, leveraging models like PhoBERT to achieve high accuracy in detecting hate speech and offensive language.

DePrBViHO/
|-- icons/                 # Folder containing icons
|-- background.js          # Background script
|-- chart.js               # Chart rendering logic
|-- content.js             # Content script
|-- dashboard.html         # HTML for the dashboard
|-- dashboard.js           # Dashboard functionality
|-- email.min.js           # Minified email library
|-- manifest.json          # Extension manifest file
|-- popup.css              # Popup styling
|-- popup.html             # Popup HTML layout
|-- popup.js               # Popup functionality
|-- styles.css             # General styling

server/
|-- ConvertModel.py        # Script to generate the phobert.onnx file
|-- main.py                # Backend server script
|-- startup.bat            # Script to start the server

1. Prepare the Model:
   Run ConvertModel.py to generate the phobert.onnx file required for the backend server.

2. Start the Backend Server:
   Use startup.bat to launch the server.

3. Enable the Extension:
   Enable Developer Mode in your browser and load the DePrBViHO extension.

Special thanks to the following resources and contributors:

• VLSP 2019
• Vietnamese Hate Speech Detection by sonlam1102
• PhoBERT Hate Speech Detection by nhattan040102
• PhoBERT Fine-tuning on Hugging Face

• NLTD2010
• leehoang0102
• And the amazing contributors who created the datasets, models, and codebases that made this project possible.