alphae-nix / BlockPhish

BlockPhish is a Google Chrome extension powered by deep learning that protects users from phishing attempts

ESIEE Paris Student project : Block Phish

Project description

This project has been made for "Day of project 2021" for ESIEE Paris. This project is composed in 2 parts :

• The first part is a web extension for chrome web browser. This extension analyse your web page, and alarm you if a phishing link is detected
• The second part is the server. It is the back part of this project. It analyses your mail, and make a prediction about your link

---

How it works ?

Start the analysis

Open your web extension and click on the "start" button. Wait few minutes for the analysis result.

Analyse

The web extension collect all links on the current page. It sends them to a python server. All links are passed in 15 functions which look at characteristics like : length of URL, using shorting services (bit.li), using weird symbol @ // #, multi subdomains, ...

Then, it returns an array of [1, 0, -1] for each link. And each array is analysed by a neural network (deep learning). It returns a percent that corresponds to a phishing prediction.

Result

This result for each link is sent to the web extension. It changes malicious link in red color on the page and an alert message appear on the web extension. For more security, malicious link are unavailable. You can go on, thanks a confirm window.

---

More about the extension

Parameters

User can change your level of risk acceptation. For example if our AI is sure at 85% that this is a phishing link would you like to identify it as phishing or not. Here we can specify from 50 to 100 %.

Name inspiration

Block Phish is a reference to stopping phishing links.

Phish is an abbreviation of phishing, but it can make you think about fish (look at our icon : it is a hook !). Block Phish also can make you think about AdBlock an extension to stop publicity pop-up.

---

Launch

Install the extension

1. Download the project
2. Go to the Chrome extension page
3. Enable developer mode
4. Click Load unpacked extension
5. Choose the extension folder and load
6. Enjoy

Install the server

1. Download the project
2. Install requirements

pip -r requierment.txt

You could face some problems with TensorFlow depending on your graphic card

3. Set Flask path

• For Windows :

cd back
set FLASK_APP = app.py

• For Linux :

cd back
export FLASK_APP = app.py

4. Run Flask server

flask run

Or :

python -m flask run

---

Demo

Careful, don't click on bad links ! It’s just for test the extension

• Facebook.com - safe link #1 ✅
• Google.fr - safe link #2 ✅
• Linkedin.com - safe link #3 ✅
• bad link #1 ❌
• bad link #2 ❌
• bad link #3 ❌

---

Contact

If you have any questions or comments about this project, or need help using it, please either raise an issue or contact us.

• Thomas Jaillon github: alias @jaillont
• Manon Hermann github: alias @CappiLucky
• Laurent Delatte github: alias @alphae-nix
• Theo Peresse-Gourbil github: alias @blackjack-nix