On the morning of June 1, 2024, the lecture on "Geospatial Entity Representation" was successfully held intheWendaoHallofWutongyu MingyiBuilding, organized jointly by the School of Computer Science andTechnology andthe Office of InternationalAffairs of HUST. Professor GaoCong, currentlyaprofessor of the School of Computer Science and Engineering at Nanyang Technological University (NTU) in Singapore and the co-director of the Singtel Cognitive and Artificial Intelligence Lab for Enterprises@NTU, delivered an excellent academic presentation on the topic of "Geospatial Entity Representation: A Step towardsCity FoundationModels."
Group photo of the guests
Professor Wei Wei addressed the opening ceremony of the lecture on behalf of the organizer. He introduced Professor GaoCong to thefaculty and students present. He hoped this lectureonacademic exchanges would benefit morepeople, promoting multidirectional communication, thinking, and learning.
Moderator Prof. Wei Wei hosts the lecture
Professor Gao Cong has worked at Aalborg University in Denmark, Microsoft Research Asia, and the University of Edinburgh. His research interests include ML4DB, spatial data management, spatiotemporal data mining, and recommendation systems. Professor Gao Cong's research papers have received the runner-up award for the best paper at the WSDM'20 and WSDM'22 conferences. He has been cited over 18,000 times on Google Scholar, with an H-index of 71. He has served as PC co-chair of ICDE'2022, vice chairman of KDD'21, PC co-chair of VLDB 2014 E&A track, and vice chairman of ICDE'18 PC. He has been the associate editor of ACM Transactions on Database Systems (TODS) and vice chairman of ACM KDD Singapore Chapter.
The lecture began with an overview of the recent research conducted by Professor Gao Cong's team, followed by a detailed introduction to the research issues of geospatial entity representation, including the representation and applications of point objects, trajectories, and regions in geospatial entities. Examples were given, such as keyword searches based on spatial relationships, point of interest recommendations, velocity inference, and regional population estimation. During the presentation, ProfessorCong provided detailed case studies, analyzing the current basic models of geospatial applications and the development trends toward city foundation models. The focus was on learning representations that facilitate the querying and analysis of geospatial entities and self-supervised learning methods applied to geospatial entities. At the end of the presentation, Professor Cong discussed the related exploratory work conducted by his team in enhancing the spatial perception capabilities of existing large models and explored possible research directions combining spatiotemporal geographic information with large models.
Prof. Gao Congis giving a lecture.
Following the presentation, Professor Cong engaged in active discussions with attendees, including students and faculty members from the School of Computer Science and Technology and the School of Architecture and Urban Planning. The focus of their discussion revolved around the integration of computational resources, geographic information, and recommendation systems within their research. This exchange of ideas contributed to a deeper understanding of the lecture among all participants.
Discussion after thepresentation
Professor Cong's detailed and captivating presentation left a profound impression on the audience, and all attendees expressed that they gained a lot from the lecture. Thislecturefacilitated academic exchanges between renowned experts, young scholars, and students, broadening the international academic perspectives ofHUST's faculty and students. It also laid a solid foundation for promoting in-depth cooperation betweenHUST and NTU.
