A multi-factor method of detecting phishing and brand impersonation sites, then logging them in a report
- Dominant Color Similarity - Using the
L*a*bColorset - Domain Cosine similarity - How similar a potential phishing domain is to a legitimate one using word embeddings
- Some whois information - such as age
- Logo comparison - Compares Logos which appear on a site to those of a site's profile
- Profiles can be found in
/src/config/company_profiles.py - Then add logos and images to the logos directory which you specify
- Use pip to install the requirements file
- CairoSVG is used for scraped SVGs from sites for logo similarity & comparison
- CairoSVG has dependencies which may depend on your Operating System. On Ubuntu Install
libcairo2 python3-dev libffi-dev ffmpeg libsm6 libxext6using the apt.
- CairoSVG has dependencies which may depend on your Operating System. On Ubuntu Install
Built as part of a Trelis Grant