An Android Malware Detection Method Using Better API Contextual Information

The vast popularity of the Android platform has fueled the rapid expansion of Android malware and existing detection methods are difficult to effectively detect malware. To address this issue, in this paper, we propose an Android malware detection method using better API contextual information (BACI). Firstly, BACI extracts the function call graph from each app. Then, we optimize the call graph by removing nodes of unknown functions while ensuring the connectivity between their predecessor and successor nodes. The optimized call graph can extract more robust API contextual information that accurately represents app behavior. Thirdly, we map the optimized call graph into a feature vector for malware detection, including three steps: call pairs extraction, call pairs abstraction, and one-hot mapping. Finally, machine learning classifiers are used for malware detection. The experimental results demonstrate that BACI greatly outperforms the existing state-of-the-art methods and can effectively detect Android malware.