Detecting Obfuscated Malware Using Graph Neural Networks

In this paper, we propose a method for detecting obfuscated malware on Android using graph neural networks (GNNs). Obfuscation is a well-known technique used by malware creators to hide themselves from being detected by anti-virus softwares. Our approach uses GNNs to represent the code of an Android app as a graph and applies graph convolutional networks to classify the app as benign or malicious. We evaluated our method on a real-world dataset of Android apps and compared it to other state-of-the-art algorithms. Our results show that our method outperformed the other algorithms in many different metrics. The proposed method has potential for application in real-world scenarios, as it can detect obfuscated malware with high accuracy.