Project Details
Unraveling the Scary Truth: How Much Your Android Applications Know About You
| Laboratory : LSIR | Semester / Master | Completed |
Description
The advances in mobile devices and technologies paved the way for the development of mobile applications that provide online services, collect sensory data, etc. Pervasiveness of localization systems on these devices is a major factor for such developments: location-based services, location-sharing applications and location- and time-stamped sensor data are becoming an everyday part of users’ lives. The downside of these applications is the threat on privacy such that based on location traces of users, it is possible to infer their interests, political orientation and personal attributes such as age and gender, which users may not want to disclose in an uncontrolled way.
In this project, we aim to develop an Android application that locally runs learning algorithms on user data to infer his/her private information (e.g., age, gender, occupation, political orientation, etc.) in order to demonstrate to users the potential privacy breaches their applications can cause. This application will read sensor data from the device (such as GPS data, accelerometer, temperature, etc.) and also other supplementary information such as calendar events for this purpose. Ultimately, the application should create awareness at end users regarding concrete privacy threats by showing them inference results such as:
- “You live in Lausanne, Switzerland.”
- “You occasionally visit Club …”
- “You are probably a student.”
- “Your political orientation is probably …”
- “You are probably a male.”
Alongside with such inference summaries, the application will also show the user which applications can infer such information based on their access permissions on the device. We believe that demonstrating to mobile device users what kind of private data is leaked in reality will enable them to grasp the risks of privacy more easily.The student is expected to design the details of the software architecture and deliver a working system at the end of the project. He will be under continuous supervision and in collaboration with the researchers. Requirements:
- Java/Android programming
- Basic database background and SQL
- Basic probability knowledge
- Machine learning knowledge is a plus
Contacts
In case of any questions, please drop us an email or come to our offices:
| Site: | |
| Contact: | Jean Paul Calbimonte Perez |