Litcius/Paper detail

Malware Family Fingerprinting Through Behavioral Analysis

Aaron Walker, Shamik Sengupta

202011 citationsDOI

Abstract

Signature-based malware detection is not always effective at detecting polymorphic variants of known malware. Malware signatures are devised to counter known threats, which also limits efficacy against new forms of malware. However, existing signatures do present the ability to classify malware based upon known malicious behavior which occurs on a victim computer. In this paper we present a method of classifying malware by family type through behavioral analysis, where the frequency of system function calls is used to fingerprint the actions of specific malware families. This in turn allows us to demonstrate a machine learning classifier which is capable of distinguishing malware by family affiliation with high accuracy.

Topics & Concepts

MalwareComputer scienceSignature (topology)Classifier (UML)Fingerprint (computing)Computer securityArtificial intelligenceSystem callBehavioral patternMachine learningMathematicsOperating systemProgramming languageGeometryAdvanced Malware Detection TechniquesNetwork Security and Intrusion DetectionDigital and Cyber Forensics