MoRePriv: Mobile OS-Wide Application Personalization

Abstract

We present MoRePriv, a system that combines the goals of privacy and content personalization in the browser. MoRePriv discovers user interests and shares them with third-parties, but only with the explicit per- mission of the user. We demonstrate how always-on user interest mining can effectively infer user interests in a mobile operating system. Inference is based on parsing and classifying multiple stream of (sensitive) informa- tion about the user, such as their email, SMS, Facebook stream, and network communications. This paper positions MoRePriv as an OS-level ser- vice that exposes information about the user in the form of a persona to applications. This provides easy-to-use APIs and limits the potential for information leaks. Us- ing a number of cases studies, we demonstrate how var- ious personalization tasks can be achieved with the help of MoRePriv.