Doctoral thesis

A privacy-aware and secure system for human memory augmentation


214 p

Thèse de doctorat: Università della Svizzera italiana, 2019

English The ubiquity of digital sensors embedded in today's mobile and wearable devices (e.g., smartphones, wearable cameras, wristbands) has made technology more intertwined with our life. Among many other things, this allows us to seamlessly log our daily experiences in increasing numbers and quality, a process known as ``lifelogging''. This practice produces a great amount of pictures and videos that can potentially improve human memory. Consider how a single photograph can bring back distant childhood memories, or how a song can help us reminisce about our last vacation. Such a vision of a ``memory augmentation system'' can offer considerable benefits, but it also raises new security and privacy challenges. Maybe obviously, a system that captures everywhere we go, and everything we say, see, and do, is greatly increasing the danger to our privacy. Any data breach of such a memory repository, whether accidental or malicious, could negatively impact both our professional and private reputation. In addition, the threat of memory manipulation might be the most worrisome aspect of a memory augmentation system: if an attacker is able to remove, add, or change our captured information, the resulting data may implant memories in our heads that never took place, or, in turn, accelerate the loss of other memories. Starting from such key challenges, this thesis investigates how to design secure memory augmentation systems. In the course of this research, we develop tools and prototypes that can be applied by researchers and system engineers to develop pervasive applications that help users capture and later recall episodic memories in a secure fashion. We build trusted sensors and protocols to securely capture and store experience data, and secure software for the secure and privacy-aware exchange of experience data with others. We explore the suitability of various access control models to put users in control of the plethora of data that the system captures on their behalf. We also explore the possibility of using in situ physical gestures to control different aspects regarding the capturing and sharing of experience data. Ultimately, this thesis contributes to the design and development of secure systems for memory augmentation.
  • English
Computer science and technology
License undefined
Persistent URL

Document views: 170 File downloads:
  • 2019INFO011.pdf: 131