We have extensive research capabilities which include:
AI has a range of ethical implications. We investigate ethical principles for AI and develop tangible frameworks for ethical AI.
We possess rich experience in data analytics skills in various domains ranging from infrastructure, energy, education, health, transport, finance to beyond.
We possess proven experience in understanding users through interpreting their observable past behaviour and adapting future interactions to their preferences, interests and needs.
We explore fairness of AI solutions from perspectives of data, model, and outcomes, as well as evaluation approaches for AI fairness.
Machine Learning is still not well understood by many real-world users. We use Human-Computer Interaction approaches to make algorithmic technologies more usable by making them more ‘Transparent’.
Human behaviour changes can be reflected from human’s multimodal responses including behaviour and physiological responses. We have solid skills in multimodal data-driven analytics to understand complex human behaviours.
A full picture of human behavioural changes in decision making is essential towards maximising the performance of human-AI teaming. We possess concrete skills in experimental design for longitudinal study for monitoring human behavioural changes.
We study human decision making – typically in AI-advised decision making systems - in order to understand and improve decision quality, user trust and confidence.
We have more than 10 years of experience with analysing data from eye-activity, galvanic skin response, EEG, mouse movements and other signals to cast light on cognitive states such as stress, emotional arousal, etc.
We investigate human trust dynamics in human-machine teaming, understand effects of both human factors and machine factors.
We are world leaders in real-time, unobtrusive measurement of mental effort via behavioural analysis and physiological sensing. Applications range from call centres to mission-critical work environments such as aviation.
Data is only half the story – we apply rigorously developed cognitive and social models to user behaviour and therefore offer predictive power – and being able to ‘nudge’ people towards desired behaviours.