AN EFFECTIVE AND EFFICIENT FEATURES VECTORS FOR RANSOMWARE DETECTION VIA MACHINE LEARNING TECHNIQUE

Authors

  • Nawaf A. Khalil Al Nahrian University
  • Ban M. Khammas Al-Nahrain University

DOI:

https://doi.org/10.31987/ijict.5.3.205

Keywords:

ransomwares detection, machine learning, cybersecurity, malware analysis, network security, feature extraction, feature selection

Abstract

Ransomware is a high major danger program that may harm any company or person and cost them hundreds of billions. Its number growing rapidly across the years.  As a result, creating a strong defense strategy against this crucial virus is required. Ransomware has grown in importance, and its consequences are becoming more severe. To solve the problem of effectively detecting ransomware, so this paper introduces a new technique to detect ransomware based on five machine learning techniques. To evaluate the proposed method, different evaluation metrics have been used.  The approach was captured n-gram characteristics based on static analysis and used n-gram vector with CF-NCF values to build the models. Using real datasets, the proposed approach shows Its ability to reliably identify between goodware and ransomware files successfully with an accuracy of classification of equal to 98.33%.

Downloads

Published

2022-12-30

How to Cite

A. Khalil, N., & M. Khammas, B. (2022). AN EFFECTIVE AND EFFICIENT FEATURES VECTORS FOR RANSOMWARE DETECTION VIA MACHINE LEARNING TECHNIQUE. Iraqi Journal of Information and Communication Technology, 5(3), 23–33. https://doi.org/10.31987/ijict.5.3.205