DrAonghus Lawlor

Dr Aonghus Lawlor’s research sits at the intersection of Artificial Intelligence, Machine Learning, and Data-Driven Personalisation, with three major pillars: Recommender Systems, AI for Sports Science, and AI for Medical Imaging.

In Recommender Systems, his work explores how to make recommender systems more adaptive, explainable, and privacy-preserving. He has developed novel approaches to enhance user trust and transparency. His research extends into federated and distributed recommender systems, allowing personalised models to be trained across multiple devices or data silos without compromising user privacy. He has also advanced reinforcement learning-based and graph neural network (GNN)-based recommenders that model complex, dynamic user-item interactions. These innovations have led to several international patents and collaborations with global industry leaders such as Samsung Research.

In AI for Sports Science, Dr Lawlor applies machine learning to large-scale athlete tracking data to optimise training and performance. He co-developed algorithms that recommend pacing strategies and training adaptations for endurance runners, aiming to enhance race outcomes and reduce injury risk. This research combines predictive modelling with behavioural analysis and was recognised with a Best Paper Award at RecSys 2019.

In Medical Imaging, his interests centre on deep learning, anomaly detection, and explainable AI. His team develops few-shot learning methods for identifying rare pathologies and interpretable models to detect and characterise disease progression—especially in Multiple Sclerosis and stroke. Through collaborations with major Dublin hospitals, he bridges computational innovation and clinical application, contributing to best paper awards and leading publications in European Radiology and Computerized Medical Imaging and Graphics.