prajapatianjali949-ui/phishing-URL-detector

GitHub: prajapatianjali949-ui/phishing-URL-detector

一个基于机器学习和 VirusTotal API 的钓鱼 URL 检测系统,解决高置信度识别恶意链接的问题。

Stars: 0 | Forks: 0

# 🛡 AI-Powered Phishing URL Detection System An intelligent cybersecurity application that detects phishing URLs using Machine Learning, advanced feature engineering, and real-time threat intelligence via VirusTotal API. ## 🚀 Project Overview Phishing attacks are one of the most common and dangerous cyber threats, where attackers trick users into visiting fake websites to steal sensitive information. This project provides a solution by combining: - Machine Learning (Random Forest) - Feature Engineering - External Threat Intelligence (VirusTotal API) to accurately classify URLs as **Legitimate** or **Phishing** with a confidence score. ## 🔍 Key Features - 🤖 AI-based URL classification - 📊 Confidence score for predictions - 🔐 Detection of phishing patterns (login, verify, secure, etc.) - 🌐 Domain analysis (TLD, subdomain, IP detection) - 🔎 URL entropy & randomness detection - 🧠 Trusted domain recognition - ⚡ VirusTotal API integration (real-time scanning) - 🎨 Clean and responsive Flask UI ## 🧠 How It Works 1. User enters a URL 2. System extracts multiple features: - URL length & structure - Special characters & digit ratio - Suspicious keywords - Entropy (randomness of URL) - Domain characteristics 3. Machine Learning model predicts: - **Legitimate (0)** - **Phishing (1)** 4. VirusTotal API checks the URL externally 5. Final result is displayed with a confidence score ## 🛠 Tech Stack - Python - Flask - Scikit-learn - Pandas & NumPy - tldextract - Requests - VirusTotal API ## 📊 Dataset The model is trained on a custom dataset containing: - ✅ Legitimate URLs (trusted domains, real-world patterns) - ⚠️ Phishing URLs (login scams, fake domains, IP-based attacks) Dataset improvements include: - Root domain variations (`google.com`, `www.google.com`) - Realistic phishing patterns (`google-login.com`) - Balanced and structured data ## 📸 Screenshots ### ✅ Legitimate URL Detection ![Legitimate](https://static.pigsec.cn/wp-content/uploads/repos/2026/04/dd9f60c18e030254.png) ### ⚠️ Phishing URL Detection ![Phishing](https://static.pigsec.cn/wp-content/uploads/repos/2026/04/23c7ac9054030256.png) ## ▶️ How to Run ### 1. Clone Repository ``` git clone https://github.com/your-username/phishing-url-detector.git cd phishing-url-detector ```
标签:AI安全, Apex, Ask搜索, Chat Copilot, Flask, NumPy, Object Callbacks, Python, Scikit-learn, Sigma 规则, SolidJS, tldextract, URL发现, URL检测, VirusTotal, 分类模型, 域名分析, 威胁情报, 安全检测, 实时检测, 开发者工具, 恶意URL, 无后门, 机器学习, 熵值检测, 特征工程, 登录欺骗, 网络安全, 网络钓鱼, 置信度评分, 逆向工具, 钓鱼检测, 随机森林, 隐私保护