https://github.com/saaaaakibhai/scamguard
This app detect the phishing website. We do use flask in backend for using our train model. To fetch the data used postman. For frontend used Flutter.
https://github.com/saaaaakibhai/scamguard
dart flask flutter joblib machine-learning pickle postman python random-forest
Last synced: 3 months ago
JSON representation
This app detect the phishing website. We do use flask in backend for using our train model. To fetch the data used postman. For frontend used Flutter.
- Host: GitHub
- URL: https://github.com/saaaaakibhai/scamguard
- Owner: Saaaaakibhai
- License: mit
- Created: 2024-12-14T10:54:46.000Z (over 1 year ago)
- Default Branch: main
- Last Pushed: 2025-01-05T13:57:24.000Z (about 1 year ago)
- Last Synced: 2025-04-01T23:32:00.083Z (12 months ago)
- Topics: dart, flask, flutter, joblib, machine-learning, pickle, postman, python, random-forest
- Language: Python
- Homepage:
- Size: 59.4 MB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: License
Awesome Lists containing this project
README
# ScamGuard
ScamGuard is a Flutter-based application that integrates a machine learning model to detect scams based on user-provided input. This project was developed as part of a university course on machine learning and serves as an example of deploying machine learning models in real-world applications.
## Features
- **Machine Learning Integration:**
- The app uses a **Random Forest Tree** model for prediction, as it provides high accuracy.
- The trained models are serialized using Python's **joblib** library to `.pkl` files for easy backend integration.
- **Backend and Frontend Communication:**
- A **Flask** backend serves as the bridge between the machine learning model and the Flutter frontend.
- The backend returns predictions in JSON format, which are then displayed in the app.
- **User Authentication (Optional):**
- Users can register and log in through the app.
- User data is securely stored in **Firebase Cloud Database**.
## Technology Stack
### Frontend:
- **Flutter:** Cross-platform framework for the mobile app.
### Backend:
- **Python:** For model integration and backend logic.
- **Flask:** Lightweight Python web framework for serving the API.
- **Firebase:** For user authentication and data storage.
### Machine Learning:
- **Random Forest Tree:** A supervised learning algorithm for predictions.
- **joblib (Python):** Used to serialize and deserialize the trained models.
## Project Workflow
1. **Train the Model:**
- The model is trained using a dataset in Python.
- Random Forest Tree is used for predictions, and the final model is exported as a `.pkl` file using **joblib**.
2. **Setup the Backend:**
- A Flask application (`app.py`) loads the `.pkl` model and serves an API endpoint for predictions.
- The backend processes user input and returns predictions in JSON format.
3. **Connect to the Frontend:**
- The Flutter app fetches data from the Flask backend via HTTP requests.
- Predictions are displayed in the app in a user-friendly format.
4. **User Authentication:**
- Optional user registration and login features are implemented using Firebase.
## Getting Started
### Prerequisites
- **Flutter Development Environment:** [Get started with Flutter](https://docs.flutter.dev/get-started/install).
- **Python Environment:** Install Python 3.7+ with Flask and joblib libraries.
### Steps to Run the Project
1. **Clone the Repository:**
```bash
git clone https://github.com/your-username/scamguard.git
cd scamguard
**pip install flask joblib**
**python app.py**
**flutter pub get**
**flutter run**
This version includes:
1. **Structured Layout**: Clear sections for features, technologies, workflow, and setup instructions.
2. **Professional Tone**: Improves readability and conveys a professional approach.
3. **Optional Sections**: Allows flexibility to include screenshots or other project details.
4. **Getting Started Instructions**: Provides step-by-step guidance for new developers.
## Here Some Demo Picture:
