CAGNet - Comprehensive Node-Labeled Graph Analysis of Mobile Malware Using Force Atlas 2 and Deep Learning

View at publisher Summarise with VIRAT
Authors: Nur Khairani Kamarudin, Ahmad Firdaus, Mohd Faris Mohd Fuzi, Nurul Hidayah Ahmad Zukri, Mohammad Hafiz bin Ismail, Muhamad Arif Hashim
Cited by: 0
References: 30

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Mobile malware continues to pose a significant threat as the use of Android-based devices expands across personal, financial, and professional domains. The constant evolution of malware techniques makes early and accurate detection a critical requirement to safeguard users. Traditional detection approaches, which rely mainly on static statistical features, often fail to capture deeper semantic patterns of malicious behavior. In this study, we propose CAGNet, a graphbased detection framework that utilizes fully labeled call graphs to preserve detailed semantic and structural information of applications. By applying the Force Atlas 2 layout to model graph relationships and employing a Convolutional Neural Network (CNN) as the detection engine, CAGNet effectively leverages node-level information for robust analysis. Experimental evaluation demonstrates that this approach achieves a highest detection accuracy of 95.89 %, surpassing conventional static analysis methods and highlighting the advantage of full node labeling in mobile malware detection.

Fields of Study & Subjects

Advanced Malware Detection Techniques Spam and Phishing Detection Security and Verification in Computing

Publication details

Journal / source
International Conference on Cognitive Systems and Computer Interaction (ICoSCI)
Pages
1–6
Publication date
15 January 2026
Language
EN