Title: Behavioural analysis and results of malware and ransomware using optimal behavioural feature set

Authors: Laxmi B. Bhagwat; Balaji M. Patil

Addresses: School of Computer Engineering and Technology, Dr. Vishwanath Karad World Peace University, Pune, India ' School of Computer Engineering and Technology, Dr. Vishwanath Karad World Peace University, Pune, India

Abstract: Ransomware is the subset of malware that is considered the most jeopardising malware. In a malware/ransomware attack, attacker encrypts all the essential data files and demands the ransom to get all the important files that it has encrypted. There are many methods or techniques, but a dynamic approach is used by many researchers to detect malware/ransomware attack. In the dynamic approach, the behavioural characteristics play a very important role in the detection of ransomware attacks. This paper presents a comprehensive analysis for the selection of the optimal behavioural feature set using various feature selection techniques. As a unique part of our research, we have found the relation and the difference between the features that can be common or different for malware and ransomware detection. We have obtained the optimal feature set for malware as well as ransomware and obtained an accuracy of 90% for XGBoost and 96.22 for KNN, respectively.

Keywords: ransomware; feature selection; malware; machine learning; Windows API calls; dynamic detection technique.

DOI: 10.1504/IJICS.2024.136719

International Journal of Information and Computer Security, 2024 Vol.23 No.1, pp.57 - 78

Received: 01 Oct 2022
Accepted: 20 Mar 2023
Published online: 19 Feb 2024 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article