PhiTector - Phishing Detector

Project Description

This project aims to detect phishing websites using machine learning techniques. We trained our model on a synthetic Kaggle dataset and deployed it on both a Streamlit web page and a Chromium extension, which makes it compatible with major browsers like Chrome, Edge, Opera, Brave etc. and would help the average consumer in identifying potential phishing websites.

Submission for CIA-2 for Foundations of Data Science, Semester 2, 2023-27 batch.

Team Phishermen 🐟🎣

• Shivsrijit
• Raghav Sridharan
• Saran Shankar
• Prajesh Raam

Tech Stack

• Pandas
• Numpy
• Matplotlib
• Scikit-Learn
• SQLite
• Streamlit
• Manifest v2 (Chromium extension)
• FastAPI

Kaggle Dataset

Requirements

Install the required modules by running the following command after cd'ing to the working folder.

pip install -r requirements.txt

Running the program

Streamlit:

Run the Streamlit file by using either of these commands:

python -m streamlit run streamlit_app.py

or

streamlit run streamlit_app.py

Chromium Extension:

1. Load the extension files to your Chromium-based browser by using the "Load unpacked extension" in settings.
2. Run main.py in the background.
3. Open a website and click on the extension's icon.
4. Once processed, the percentage of legitimacy of the website will be displayed.