Home // CYBER 2016, The First International Conference on Cyber-Technologies and Cyber-Systems // View article

Quantifiable Measurement Scheme for Mobile Code Vulnerability Based on Machine-Learned API Features

Authors:
Hyunki Kim
Joonsang Yoo
Jeong Hyun Yi

Keywords: Android; Vulnerability; Application Assessment

Abstract:
Owing to open market policies and self-signed certificates, any malicious application developer can easily insert malicious code into Android mobile applications and then distribute them in the Google Play market. Furthermore, even applications that are known to be benign or safe are collecting private information without asking users. Thus, there is a need for a quantifiable measurement scheme that can evaluate the degree of risk posed by an application beyond applications simply being classified as normal or malicious. In this paper, by using ensemble learning, we develop a quantifiable measurement scheme to assess the sensitivity of the Android framework API, and we experimentally evaluate the feasibility of this scheme.

Pages: 27 to 28

Copyright: Copyright (c) IARIA, 2016

Publication date: October 9, 2016

Published in: conference

ISSN: 2519-8599

ISBN: 978-1-61208-512-8

Location: Venice, Italy

Dates: from October 9, 2016 to October 13, 2016