This repository contains notebook demos on encoding strings as images using the Word2Vec model and performing predictions on those images using ML techniques. I recently came across a reasearch paper which takes a similar approach for detecting malware. We can probably apply some of the concepts they worked on to this problem space. Check out the notebooks/POC folder to see how this would work.

https://www.intel.com/content/www/us/en/artificial-intelligence/documents/stamina-deep-learning-for-malware-protection-whitepaper.html

1. Clone this repository
2. pip install git https://github.com/tteigman/string2image

• Data cleansing by detecting strings in a data field that are out of place (and putting them in place)
• Domain specific fuzzy string matching