PhishNet
PhishNet – Spot the Scam Before It Strikes !
Created on 28th May 2025
•
PhishNet
PhishNet – Spot the Scam Before It Strikes !
The problem PhishNet solves
Cybercrime is no longer a distant threat—it’s at our doorstep. Every day, thousands of people unknowingly click on malicious links or fall for fake SMS messages. Most of these attacks rely on one simple weakness: lack of awareness.
Phishing tools like Zphisher are making it easier than ever for attackers to replicate real websites and steal personal information. Users, especially non-tech-savvy ones, often have no idea how to spot a fake link or a scam message.
PhishNet Cyber Guardian solves this by giving users a smart, interactive cybersecurity assistant they can trust. It offers:
🔗 Phishing URL Detection: Instantly analyzes links and gives a trust score with explanations.
📩 SMS Scam Detection: Users can upload text or image-based SMS messages to detect scams using OCR and machine learning.
🧠 Conversational AI Interface: A chatbot-style system that makes cybersecurity guidance simple, friendly, and accessible.
🖼️ Live URL Preview & Source Viewer: Lets users see what a URL would load without actually visiting it.
🌐 Visual Threat Feedback: Color-coded feedback and alerts to help even non-technical users understand the threat.
By turning security into a conversation, not a complication, PhishNet makes the digital world safer, one message at a time.
Challenges we ran into
We ran into several interesting challenges while developing PhishNet:
🕵️♂️ Detecting Real-World Phishing Links :
Many phishing kits replicate login pages perfectly. Detecting them required more than just keyword checks. We built heuristic-based filters and analyzed URL structures, metadata, and hosting behaviors to identify suspicious links with higher accuracy.
🧾 Extracting Scam Text from SMS Images:
Integrating Tesseract OCR for scam SMS detection was tricky. Poor image quality or blurry screenshots led to inaccurate text extraction. We tackled this by adding image preprocessing steps like binarization, noise removal, and contrast enhancement.
💬 Creating a Seamless Chatbot UI:
Making the chatbot interactive, fast, and visually intuitive was a UX challenge. We used JavaScript and styled components to simulate a real-time chat experience with typing indicators, button suggestions, and animated responses.
Tracks Applied (1)
Ethereum Track
ETHIndia