A tool for visual analysis of permission-based data access on Android phones

Abstract: Privacy is a topic of ever-increasing interest in the modern, connected world. With the advent of smart phones, the boundary between internet and the personal sphere has become less distinct. Most users of smart phones have very vague ideas of how various apps intrude on their privacy. At Karlstad University, one of the research groups at the Department of Computer Science is dedicated to research on privacy and protection thereof in general, and one research project is about increasing knowledge of how apps collect personal information about its users. This master thesis is about the development of a visualization tool that processes data collected from Android devices by a surveillance app developed within the aforementioned research project. The app keeps record of the usage of what in Android is termed Dangerous Permissions. The info collected is when such an event occurs, which app has requested the permission, the permission requested as well as the geographical location at the time of the event. Over time, more than 2 million such events has been recorded and collected in this manner. Previously, two student projects have developed different web based tools for visualizing the data collected. With this thesis work, a desktop application was developed; a visualization tool that imports the aforementioned data into a database connected to the tool. The graphical user interface of the visualization tool allows an analyst or scientific researcher to do detailed and fine-grained searches in that database and present the result in various charts, thereby visualizing how the apps ask for information that can be used for identifying and surveying individuals, thereby intruding on their privacy. The visualization tool is carefully designed with the aim for it to be scalable and extendable, through an architecture that allows for continuous development of further visualizations as well as other analysis functionality.