PhishWatch

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.

Instructions

Here are the instructions to run our project.

Running our model

import necessary libraries like numpy, pandas, matplotlib, sckit-learn

pip install -r requirements.txt

running the frontend

go to frontend directory using command

cd frontend

install the node_modules and run the development server

npm install
npm run dev