Blocking in the Internet of Things
It has been shown that many mobile apps, and IoT devices communicate to a range of servers, some of which appear beyond that needed to function. This raises privacy considerations, given that the sensors and actuators deployed in our homes, streets, carried in our pockets, etc, can generate much sensitive information. This project involves undertaking a technical survey, to investigate the extent to which seemingly spurious traffic can be blocked, while still enabling appropriate functioning of the device/application.
ML explainability: a systematic comparison
There is much discussion regarding making ML models more explainable. While there is intuition that certain model building methods are explainable than others - e.g. neural networks vs. decision trees - there is little in the way of a rigorous comparison between the approaches. This project is to perform such a comparison, by defining various metrics, criteria, and systematically testing a range of approaches, to provide data indicating the differences and relevant factors/considerations.
Facilitating user-centric system interrogation
There are increased demands for greater transparency regarding online services, particularly where ‘users’ are subject to ‘algorithmic’ decisions. Though the providers of services have detailed information, e.g. regarding the run-time configuration, software/model versioning, etc, much of this is aimed at systems operation, rather than user concerns. Moreover, responses given to a user seeking an explanation of systems may not always be accurate, given that systems are often updated, reconfigured and may otherwise change. This project is to devise mechanisms that facilitate a user-centric systems audit, through means that can connect a user’s interaction with a system to the particulars of the run-time environment, and perhaps also the design-time considerations.
Attacking ML models through explanations
There is much value in machine learning models, both in terms of the quality of their outputs, and by way of the data on which they are built. At the same time, we are seeing increased demands for greater transparency regarding ML models, including mechanisms for describing why particular decisions were made. There is, however, some potential for model explanations to introduce security and privacy risks, e.g. by facilitating an attacker in uncovering pertinent aspects of the model, or indeed, uncovering details of the training data. As such, this project will explore the propensity for model transparency mechanisms to enable such attacks.
Enabling accountable automated decision-making
Many of the concerns regarding automated decision-making relates to the processes and context in which the models are built, deployed and operated. Towards this, decision provenance is a concept that concerns tracking the flow of data to assist accountability concerns. This project involves taking forward the concept, implementing and exploring it in various ML scenarios.
Supporting Seamlessless: A reconfigurable, command-and-control framework for the Internet of Things
Visions for future computing environments are of a hyperconnected world, where a bunch of sensors, actuators, and other components to provide functionality seamlessly. A rather distopian view of such a world is illustrated in Minority Report. Such an environment requires a flexible management infrastructure 'under-the-hood'. This project involves designing and engineering a management infrastruture to enable such functionality, aiming at supporting an IoT environment. It has a strong practical/engineering component, the aim being for an open source release.
What's happening on social media?
There is much concern over the content being disseminated social media networks - misinformation and fake news, political targetting, hate speech, illegal content, etc. Currently, most platforms require a user to report problematic content, which is then manually assessed and removed. This project is to explore and develop alternative approaches, considering what technical signals might indicate to a social media platform, or indeed, to users, the content that might warrant extra scruitny. This would involve considering patterns of sharing, posting, external linkage, etc.
Physical world privacy (`Do Not Track' in a smart city)
Data being collected constantly from everywhere - environmental sensors, cameras, etc. A lot of this through relating to the physical environment/space you're in, and those around you. At present there are few means, technical or otherwise, for opting out. This project is to explore means to have some control over the data collected on you within a physical space. Imagine the ability to 'tell' someone's Alexa 'not to hear me', or a Google-glass wearer to `ignore me'.
Biohacking: security, privacy and safety
There is much excitement with promise of smart prosthetics, microchips for monitoring health and releasing medication, and other physical body augmentation by way of implants (e.g. bio-hacking). While the technology is in its infancy, it is expected to gain significant traction in the coming decade. However, the interconnection between computer networks and the human body present a number of privacy, security and safety risks at very personal level.
This project seeks to develop an understanding of how security and privacy mechanisms can be used to protect users from technical failures, the misuse or interference of "bio-devices" from parties with malicious intent, and to explore potential ways forward.
Physical world gaming
Pokemon-go brought the concept of augmented reality-driven gaming to the masses. However, many games in the space tend to be location and image based -- i.e. responding to a person's (device's) location, and viewpoint. However, as the Internet of Things evolves, where connected 'things' have actuation capabilities, there is real potential for gaming to bring about physical-world effects. This project is to explore the potential in this space')'
Blockchain and beyond!
There is a lot of excitement around blockchain to enable a whole range of new systems and services. But we must work through the hype! There are many possible projects in this space: from managing micropayments in 'connected worlds' (smart-cities, IoT); security/vulnerability analysis; using blockchain to tackle 'fake news', etc. Do get in contact to discuss
Playing with movement
Most current-gen mobile devices have a bunch of accelerometers, gyroscropes, and other related sensor streams that can be used for various purposes.
This is an experimental-based project basically about doing more with movement.
For example, can certain movements (or patterns there of) uniquely identify someone? How easy is it to detect when groups of people are peforming similar movements, using different devices etc. Lots of cool applications for this - everything from security (phone locks, PKI) to enabling collaboration between people in gaming or event spaces