Phishing Website Detection Browser Extension with Machine Learning
A Chrome extension that warns users before they open a phishing site, using an ML model trained on URL and page features.
Representative screens. Your project is built to the listed features after you order.
Need your guide to approve it first?
Download the two-page synopsis and abstract, free.
Compiled for you, not resold
Every codebase is compiled to order, configured with its own database migrations and bundled with a report written to your university’s format, so your submission never matches anyone else’s. Deliverables reach your vault within 10 to 12 working days, depending on demand and how much customisation you ask for.
Problem Statement & Abstract
Phishing remains the most common way accounts and money are stolen. This project protects users at the moment it matters: a Chrome extension checks every page the user visits and shows a clear warning before they type a password into a fake site.
The extension extracts features from the URL (length, number of dots and hyphens, use of an IP address, suspicious words, HTTPS, domain age) and from the page (forms posting to another domain, hidden iframes, external favicons). A FastAPI service runs a Random Forest model trained on a public phishing dataset of about 11,000 sites and returns a risk score in milliseconds; well-known safe domains are allow-listed and results are cached.
High-risk pages are blocked with a warning screen that explains why, while medium-risk pages show a banner. Users can report false positives, which are logged for retraining. The report compares Logistic Regression, Random Forest and XGBoost with precision, recall and F1, and discusses why recall matters most for security tools.
Key Features
- Real-time check of every visited page
- Over 20 URL and page features extracted in the browser
- Random Forest model served by a FastAPI API
- Block page with the reasons for high-risk sites
- Warning banner for medium-risk sites
- Allow-list and caching for speed
- False-positive reporting for retraining
- Model comparison with precision, recall and F1
Suitable For
Everything included
+ GST • about $107 USD
Added to your cart
- Delivered in 10 to 12 working days
- Free revisions for 30 days while your guide reviews
- Report written to your university’s format