Towards Scalable Evaluation of Mobile Applications through Crowdsourcing and Automation (CMU-CyLab-12-006)

Shahriyar Amini, Carnegie Mellon University
Jialiu Lin, Carnegie Mellon University
Jason Hong, Carnegie Mellon University
Joy Zhang, Carnegie Mellon University
Janne Lindqvist, Rutgers University

Abstract or Table of Contents

With the widespread adoption of smartphones, mobile applications have gained mainstream popularity. However, the potential privacy and security risks associated with using mobile apps are quite high, as smartphones become increasingly integrated with our lives, being able to access our email, social networking accounts, financial information, personal photos, and even our cars and homes. To address this problem, we introduce AppScanner, an automated cloud-based service based on crowdsourcing and traditional security approaches to analyze mobile applications. Considering the large and growing number of mobile applications, our envisioned service builds on crowdsourcing, virtualization, and automation to enable large-scale analysis of apps. AppScanner provides end-users with more understandable information regarding what mobile apps are really doing on their devices. This paper offers an overview of our vision for building AppScanner, as well as work to date in specific components, including automated traversal and monitoring of mobile applications, and interactive visual presentation of app traversal results. Armed with transparent and descriptive information regarding app behavior, users can make better decisions when installing and running apps.