DrAonghus Lawlor

Dr Aonghus Lawlor is an Assistant Professor in the School of Computer Science at University College Dublin (UCD) and a Funded Investigator at the Insight Centre for Data Analytics.

His research lies at the intersection of Recommender Systems, AI for Sports Science, and Medical Imaging, where he has established research directions spanning fundamental machine-learning, applied AI and industrial collaboration.

In the field of Recommender Systems, Dr Lawlor has advanced the use of distributed and generative recommendation models, contributing early work (2018) on generative text models for transparency and user trust. His group has developed distributed and federated recommender systems designed to preserve data privacy while supporting large-scale, rapid model training. These innovations underpin a long-term collaboration with Samsung Research (South Korea), leading to multiple international patents. His team has also pioneered reinforcement-learning recommender systems and graph-based models, extending the theoretical foundations of graph spectral processing and graph neural networks (GNNs) for personalisation.

In AI for Sports Science, Dr Lawlor works closely with Prof. Barry Smyth to analyse data from marathon runners, with the goal of improving training effectiveness and injury prevention. Using large-scale exercise-tracking data, they designed algorithms to recommend optimal pacing strategies—a project that earned a Best Paper Award at RecSys 2019.

In Medical Imaging, he leads interdisciplinary projects with clinicians at St Vincent’s and the Mater Hospital Dublin, developing few-shot anomaly detection and explainable deep-learning methods for radiology. His team won a Best Paper Award at MIUA 2021 for their work on few-shot learning for anomaly detection, and recent collaborations have yielded high-impact papers in Computerized Medical Imaging and Graphics (2024), European Radiology, and the American Journal of Neuroradiology. Current projects explore progression modelling in Multiple Sclerosis and interpretable ML for stroke prognosis.

Dr Lawlor contributes extensively to the international research community as a programme-committee member and reviewer for leading conferences such as ACM RecSys, SIGIR, ECIR, WWW, AAI, IJCAI among others.

Google Scholar

ORCID