Professor Georg Gottlob awarded Royal Society Research Professorship
These prestigious appointments provide long-term support for internationally recognised scientists of exceptional accomplishments from a range of diverse fields, including biochemistry, genetics, mathematics, chemistry, computer science, developmental biology and physics.
The Research Professorships help to release these exceptional researchers from competing duties, such as teaching and administration, allowing them to focus on ambitious and original research of the highest quality. The awards also enable distinguished, international research talent to relocate to a UK academic institution.
Venki Ramakrishnan, President of the Royal Society, said: ‘We are delighted with the six appointments made in this year’s Royal Society Research Professorship competition.
'The newly appointed Research Professors join a world-class cohort of leading scientists that have and continue to make exceptional contributions to science. This type of investment in world-leading talent is crucial to the continued success of UK science’.
Professor Gottlob’s research project is entitled ‘RAISON DATA – Rule-based AI Systems for Reasoning on Massive Data’. He is working to build better, and more trustworthy, artificial intelligence (AI) by combining existing machine learning approaches with the use of transferable knowledge that humans deploy.
While machine learning has advanced rapidly thanks to neural networks, the trial and error approach it uses is similar to the way human babies learn to walk or recognise their parents. But trial and error learning can be time consuming and risky, for example, in the case of identifying what foods are poisonous. It is sometimes also impossible, due to the lack of data. However, humans developed language and other systems which make it possible to pass on knowledge directly to others who integrate that into their own knowledge. Professor Gottlob’s professorship will look at the rules which enable knowledge transfer to work best in AI and integrate them with existing machine learning approaches. He will then build software that aims to apply this reasoning-based learning in useful, real-life situations.
Find out more about Royal Society Research Professorships here.