Reproduce “PhishSnap: Image-Based Phishing Detection Using Perceptual Hashing” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Md Abdul Ahad Minhaz, Zannatul Zahan Meem, Md. Shohrab Hossain
Research from United International University introduces PhishSnap, an on-device system for image-based phishing detection that utilizes perceptual hashing (pHash) to identify visually deceptive webpages. The browser extension achieves real-time, local processing of webpage screenshots for privacy-preserving analysis, demonstrating an accuracy of 0.79 in distinguishing phishing attempts.
Key insights
- Perceptual hashing provides a lightweight and effective method for quantifying visual similarity between webpages, suitable for on-device phishing detection.
- Performing all detection processes locally within a browser extension eliminates the need for data transmission to remote servers, addressing critical user privacy and latency concerns.
- The diversity and quantity of available phishing screenshots in datasets, often limited by rapid takedowns, can impact the generalizability and reported accuracy of visual detection systems.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.