About This Project

What is PhishGuard?

PhishGuard URL Phishing Detection System is an AI-powered web application designed to protect users from phishing websites. It uses machine learning to analyze URLs in real-time and classify them as safe or malicious.

Built as a full-stack project combining Flask (Python), scikit-learn for ML, and a modern glassmorphic UI. The system performs WHOIS lookups, DNS resolution, and extracts 12+ features from each URL to make accurate predictions.

Python Flask scikit-learn SQLite JavaScript Bootstrap 5
Project Collaboration

PhishGuard is proudly developed as an MCA Final Year Project, representing a collaborative innovation in cybersecurity and machine learning. This system reflects the combined technical expertise, research effort, and full-stack development skills of:

Nikit Singh Kanyal
AI Model Development, Backend Architecture & System Integration
Vaibhav Vishwakarma
Frontend Engineering, Database Design & Application Deployment
Key Capabilities
Real-time ML Detection
Trained Random Forest model classifies URLs with high accuracy and provides confidence scores.
Domain WHOIS Analysis
Checks domain registration age, registrar, creation and expiry dates for suspicious patterns.
User Dashboard & History
Personal scan history, detection stats, and exportable data for every registered user.
Admin Control Panel
User management, system logs, role-based access, and ability to register new users.
Advanced AI Protection

Secure Your Digital Frontier.

This system uses state-of-the-art machine learning to analyze URLs in real-time, protecting you from sophisticated phishing attacks.

https://

Multi-Layered Intelligence

Our engine looks at over 12 distinct signals to determine risk.

AI Pattern Matching

Random Forest model detects subtle URL obfuscation techniques that bypass traditional filters.

WHOIS Intelligence

Real-time domain age and registrar verification to spot newly registered malicious domains.

Redirect Resolution

Follows URL shorteners and complex redirect chains to analyze the final destination.

Scan History

Every scan is logged in your personal dashboard so you can review past checks anytime.

Step-by-Step Process

How It Works

From paste to verdict in seconds. Here's what happens behind the scenes when you scan a URL.

1
Paste Your URL

Enter any suspicious link into the scanner. We auto-detect the protocol and www prefix.

2
Feature Extraction

We extract 12+ features: URL length, special chars, digit ratio, HTTPS status, subdomain count, and more.

3
AI Analysis

A trained Random Forest model with 100 decision trees votes on whether the URL is phishing or safe.

4
Instant Verdict

Get a clear Safe/Phishing verdict with confidence score, domain WHOIS info, and safety recommendations.

12+

URL Features Analyzed

100

Decision Trees Voting

2s

Average Scan Time

WHOIS

Domain Intelligence