AI & Big Data

• Mission Priority Areas
• Supervisors
• Partners
• Your Favourites

Click / tap the stars next to items in the CoP to mark your favourites.

Glasgow’s AI research is uniquely focused on developing approaches that are human-centred and user-oriented. This means deploying approaches that understand how humans communicate and interact with the AI system, and how to make it blend smoothly with our daily life and social interactions.

The UKRI Centre for Doctoral Training in Socially Intelligent Artificial Agents works on endowing machines with social intelligence, i.e., with the ability to understand social and psychological phenomena such as emotions, personality, attitudes, intentions, etc. (see Social AI Glasgow).

Natural Language Processing activities target AI approaches that make sense of language, whether this means to provide people with personalised financial advice (see INFINITECH, a project led by Prof. Craig Macdonald, Dr Richard McCreadie and Prof. Iadh Ounis, and FAR-AI), to support the interaction with large and complex data sets (see Closed-Loop Data Science) or to elicit user feedback for improving the quality of conversational search systems (see Glasgow Information Retrieval Group), thus expanding humans’ cognition and skills in dealing with large amounts of information.

Finally, Human Computer Interaction activities investigate the AI as a means to support the interaction between people and machines, whether these are humanoid robots (see MuMMER), virtual environments populated with artificial humans (see SONICOM) or people interacting in Extended Reality (see FUSION).

All work above requires not only technical and computing skills, but also consistent attention to ethical and societal issues. Hence, we actively contribute to the Responsible AI Hub and its mission trough the development of approaches capable to be fair with their users, to monitor fair representation of people in media or to ensure conducting research on medical records while preserving the patients’ privacy (see SIEPH: Safe Information Extraction from Patient Histories).