Enhancing Malware Detection Using 'Genetic Markers' and Machine Learning

Publication Date

1-1-2023

Document Type

Conference Proceeding

Publication Title

2023 IEEE International Conference on Dependable, Autonomic and Secure Computing, International Conference on Pervasive Intelligence and Computing, International Conference on Cloud and Big Data Computing, International Conference on Cyber Science and Technology Congress, DASC/PiCom/CBDCom/CyberSciTech 2023

DOI

10.1109/DASC/PiCom/CBDCom/Cy59711.2023.10361372

First Page

975

Last Page

977

Abstract

Despite the advanced techniques of malware detection using machine learning and deep learning, our community has been still suffering from new variants of malware in networks and systems. This paper proposes a new NLP-based malware detection method to generate a genetic marker by capturing the semantic behaviors of each malware family. The unique genetic markers aim to identify a specific malware family since each malware family has different characteristics of the distribution of opcodes. This paper evaluates the malware behavior based on opcodes for each family by using an NLP model and creates different templates to identify each malware family. Based on our experiments, our new approach achieved more than 99% detection rates and is fault-tolerant against various malware obfuscation techniques since the method captures the meaningful context of each malware sample.

Keywords

Deep Learning, Machine Learning, Malware Detection, Natural Language Processing

Department

Computer Science

Link to Full Text

Share

COinS