PhishWatch is a website that leverages machine learning to determine whether a URL is phishing or legitimate. Our machine learning model is trained on a dataset from https://github.com/GregaVrbancic/Phishing-Dataset, which includes over 80,000 URLs with 111 features. We utilize the K-Nearest Neighbors classification algorithm to achieve optimal results. The frontend is built with React.js, and FastAPI is used for API endpoints that extracts features from the given url and predict it using our pre-trained model.
Here are the instructions to run our project.
import necessary libraries like numpy, pandas, matplotlib, sckit-learn
pip install -r requirements.txtgo to frontend directory using command
cd frontendinstall the node_modules and run the development server
npm install
npm run dev