An Android Malware Detection Method Based on Feature Codes

DOI

10.2991/icmmcce-15.2015.519How to use a DOI?

Keywords

android security, malware, feature codes, system call, machine learning.

Abstract

The Linux-based android operation system is now exposed to high risks of security since the malware of smart phone explodes. For the purpose of effectively detecting the malware on the android platform, an android malware detection method based on feature codes is described in this paper. By using the function call and system call, analyzed and extracted from the malware sample library, as the feature vectors which will be subject to training and classification upon machine learning and data mining algorithm, a feature library and a detection model is established. An android malware detection system, ANDect, is developed upon this method and used for detecting 350 malicious applications and 750 non-malicious applications. As the results, ANDect is proven that it can effectively find out the undiscovered malicious Applications of android by utilizing the feature vectors of codes from the android applications, with high accuracy and low false positive rate.

Copyright

© 2015, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Yiran Li
AU  - Zhengping Jin
PY  - 2015/12
DA  - 2015/12
TI  - An Android Malware Detection Method Based on Feature Codes
BT  - Proceedings of the 4th International Conference on Mechatronics, Materials, Chemistry and Computer Engineering 2015
PB  - Atlantis Press
SN  - 2352-538X
UR  - https://doi.org/10.2991/icmmcce-15.2015.519
DO  - 10.2991/icmmcce-15.2015.519
ID  - Li2015/12
ER  -