Modern information processing tasks often involve data that come with not only a large volume but also increasingly complex structures. In particular, an increasing amount of real-world data are being collected in topologically-complicated domains, such as social and biomedical networks, with inherent relationships behind the observations.
To cope with challenges that come with such complexities, we utilise graphs as mathematical tools to model relationships and structures in the data, with a particular interest in the fast-growing fields of graph signal processing and geometric deep learning.
Topics of interest:
Seminars/Reading groups:
By utilising digital tools through the unique lens of "Big Data", computational social science studies human and social behaviour in the modern society and represents a paradigm shift in social sciences research.
At the heart of computational social science are networks formed by people within the physical and social environment, which can be conveniently modelled by a variety of graph representations. We are thus interested in utilising graph-based data processing techniques to bring unique contributions to this new field of studies.
Topics of interest: