The Research Chairs and Senior Research Fellowships appointments include the Department of Computing's researcher Professor Francesca Toni.
The Royal Academy of Engineering has today announced the appointment of four new Research Chairs and two Senior Research Fellows at universities across the UK who will spend the next five years working with some of the world’s largest companies to tackle a broad spectrum of engineering challenges.
The Research Chairs and Senior Research Fellowships appointments include the Department of Computing researcher Professor Francesca Toni who is working alongside J.P. Morgan, focusing on developing forms of interactive explainable AI that can help explain how AI systems work by providing justifications for system decisions.
She says: "The Research Chair will provide a unique opportunity to bring together the strengths of the Computational Logic and Argumentation Group in the Department of Computing at Imperial College London, J.P. Morgan and the Royal Academy of Engineering, to address explainability challenges for AI in general and particularly in the financial sector".
It is widely acknowledged that the difficulty of explaining AI is a barrier to uptake. Existing efforts towards explainable AI envisage static ‘one-way’ (machine to human) interactions. This Research Chair will deliver machines, powered by a variety of AI methods, that can engage with humans in two-way explanatory dialogues. The machines will explain their recommendations and humans will question their explanations and provide feedback. Such machines will be able to work in synergy with humans within a human-centred but AI-supported society.
This ambitious research vision will be realised using computational argumentation based on symbolic AI. Professor Toni will define argumentative abstractions for a variety of AI methods from which various types of explanations can be drawn to generate argumentative dialogues between humans and machines.
Article text (excluding photos or graphics) © Imperial College London.
Photos and graphics subject to third party copyright used with permission or © Imperial College London.
Mr Ahmed Idle
Department of Computing