Adobe Systems recently released an open source tool to help security pros in the fight against malware.
According to the company, the Adobe Malware Classifier is a command-line tool aimed to help antivirus analysts, IT administrators and security researchers determine if a binary file contains malware so they can develop malware detection signatures faster. The tool uses machine-learning algorithms to categorize Win32 binaries – exe and DLL files – into three classes: 0, representing clean; 1, representing malicious; and unknown.
“Part of what we do at PSIRT is respond to security incidents. Sometimes this involves analyzing malware,” blogged Karthik Raman, a security engineer at Adobe. “To make life easier, I wrote a Python tool for quick malware triage for our team. I’ve since decided to make this tool, called “Adobe Malware Classifier,” available to other first responders (malware analysts, IT admins and security researchers of any stripe) as an open-source tool, since you might find it equally helpful.”
The tool extracts seven key attributes from every analyzed binary file and feed them to one or all four of the classifiers and presents the results. It was developed using models resulting from running the J48, J48 Graft, PART and Ridor machine-learning algorithms on a data set of approximately 100,000 malicious programs and 16,000 clean programs, Raman explained.
The tool, which has been met with mixed reviews on Reddit.com, is available here.
