Distinguished Professor Philip S. Yu Explores Geometric Deep Graph Learning through a New Lens
10 Dec 2024
Professor Yu Sheds light on Riemannian Geometric Deep Graph Learning in building complexed graph foundation models.
The lecture draws a full house of participants.
Professor Michael Ng, Dean of Science (right), and Professor Jianliang Xu, Head of the Department (left), present the souvenir to Professor Yu.
Distinguished Professor Philip S. Yu, a prominent figure in the field of mining, fusion and anonymization of big data from the University of Illinois Chicago, delivered an insightful lecture titled “Geometric Deep Graph Learning: A New Perspective on Graph Foundation Model” on 6 December 2024.
In this lecture, Professor Yu explored the limitations of deep graph learning in traditional Euclidean space, emphasising the necessity for more effective representation spaces to model complex graph structures comprehensively. He pointed out that existing graph models in Euclidean space often oversimplify the unique geometries of complex graphs. To address this issue, he explored Riemannian geometric approach in Deep Graph Learning and introduced the application of Mixed Curvature Riemannian Space Representation, which effectively quantifies graph structure and geometry. His findings demonstrated that the Riemannian Geometric Deep Graph Learning approach not only outperforms traditional methods in classification and link prediction but also effectively models irregular clusters and dynamic networks. Professor Yu’s insights into Geometric Deep Graph Learning illuminate a new direction for developing graph foundation models for complex graphs.
Professor Philip S. Yu is a Distinguished Professor and the Wexler Chair in Information Technology at the Department of Computer Science, University of Illinois Chicago. He is the Fellow of the ACM and IEEE and has received numerous accolades for his impactful research in big data mining, fusion, and anonymization. Professor Yu is the recipient of ACM SIGKDD 2016 Innovation Award, the IEEE Computer Society’s 2013 Technical Achievement Award and the Research Contributions Award from ICDM in 2003 for his pioneering contributions to the field of data mining.
