Android, the world’s most widely used mobile operating system, is the target of a large number of malwares. These malwares have brought great trouble to information security and users’ privacy, such as leaking personal information, secretly downloading programs to consume data, and secretly sending deduction SMS messages. With the increase of malwares, detection methods have been proposed constantly. Especially in recent years, the malware detection methods based on deep learning are popular. However, the detection methods based on static features have a low accuracy, and others based on dynamic features take a long time, all this limits its scope. In this paper, we proposed a static feature detection method based on deep learning. It extracts specific API calls of applications and uses DNN network for detection. With the dataset composed about 4000 applications and extremely short time, it can achieve an accuracy rate of more than 99%.
CITATION STYLE
Jin, Y., Yang, T., Li, Y., & Xie, H. (2020). Effective Android Malware Detection Based on Deep Learning. In Communications in Computer and Information Science (Vol. 1252 CCIS, pp. 206–218). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-981-15-8083-3_19
Mendeley helps you to discover research relevant for your work.