Bruce W Wilson
Decisions made by machine learning algorithms impact huge amounts of the population. These individual stakeholders can range from data scientists to end-users, however they all can ask the same question: why this decision? Explainable Artificial Intelligence (XAI) aims to produce explanations to these decisions, but with the wide range of impacted stakeholders, how explanations are presented to and digested by the user can differ greatly. This report intends to evaluate explanation generation methods and how they can be used to impact presentation methods. Alongside this, it reviews analysis and multi-modal data fusion methods of various user-state features with relation to reducing frustration and cognitive overload. Finally, these topics are combined to enable the creation of an adaptive information presentation system that reduces cognitive overload and can apply to a variety of situations or stakeholders.
I graduated from Heriot-Watt University with a distinction in Masters of Engineering in Software Engineering, alongside winning the British Computer Society Prize for best student on the MEng course.
In my spare time I work as an ambassador for the Scottish Huntington’s Association and HDYO, and proudly serve as an HD-Community Advisory Board team member.
