MAIN RESEARCH AREAS

Within AI Theme, there is extensive expertise in machine learning and deep learning, natural language processing, computer vision, knowledge engineering, planning and scheduling, AI ethics, and computer systems for AI. Applications include:

• Health: Biomarker/target discovery, e.g., virus detection, drug discovery; Human-centred medical imaging, e.g., explainable AI for age-related macular degeneration; Hypothesis generation, e.g., WHEEL intervention (in collaboration with Prof Nathan Congdon)
• Fintech: Anomaly/outlier discovery, e.g., cyber security; Compliance audit, within Advanced Research and Engineering Centre with PwC;
• Media: Semantic search (to search based on semantic embedding of content), e.g. multimodal video search
• Others: Time tabling; AI for education e.g., question/answer generation

There is a growing body of work on Discovery AI, which is a combination of AI techniques to facilitate discovery research including causality discovery, model discovery and hypothesis generation. DAI aims to establish causal, correlational and logical relationships between entities based on different types of data including observational data, experimental data, ontologies, and scientific literature. 

• Deep learning: to represent data semantically for purposes (e.g. classification) and to build models and to characterise interested classes of data.
• Knowledge and reasoning: to extract and represent knowledge, to draw conclusions, and to generalise to generate new hypotheses.
• Causality and Explainability: to discover causal relationships based on observational data and explain models and predictions on the basis of causal relationships

OUR IMPACT

Artificial Intelligence Research in Northern Ireland

SPINOUT COMPANIES

• Anyvision 

• EventMAP  

• Brainwavebank

• Analytics Engines