SeGDroid: An Android malware detection method based on sensitive function call graph learning

Zhen Liu, Ruoyu Wang, Nathalie Japkowicz, Heitor Murilo Gomes, Bitao Peng, Wenbin Zhang

Abstract: Highlights•We devise a novel Android malware detection framework.•We propose a graph pruning method that preserves the context of sensitive APIs.•We propose a node representation method for learning semantic knowledge of APIs and opcodes.•We propose a node importance calculation method for understanding mobile malicious behavior.•Our method achieves an F-score of 98% for malware detection.

External IDs:dblp:journals/eswa/LiuWJGPZ24