Wilson Wongso completes USA research trip
Author ADM+S Centre
Date 16 December 2025
ADM+S researcher Wilson Wongso, a PhD student from UNSW, has recently completed a research trip to the University of California, Berkeley and the University of Minnesota in the United States. While there, he attended a conference and delivered presentations about his research on Large Language Models (LLM’s).
Wilson attended the ACM International Conference on Advances in Geographic Information Systems (SIGSPATIAL 2025), hosted by the University of Minnesota. He travelled with fellow researchers from Collaborative Human-Centric AI Systems (CRUISE), a UNSW based research group which includes ADM+S Chief Investigator Flora Salim. Wilson also met with ADM+S Research Fellow Yufan (Tina) Kang from Monash University.
SIGSPATIAL 2025 attracted participants from a wide range of universities, institutes and industry partners, including Google and Amazon. While at the conference, Wilson presented his research on GenUP as a lightning talk and during the poster session.
“The core idea of GenUP is to generate user profiles that inform POI recommender systems, giving end-users more control over their recommendations.” Wilson said.
At SIGSPATIAL, Wilson served as a program committee member for an UrbanAI workshop, organised by Oak Ridge National Laboratory, UNSW and Emory University. He was also present for supervisor Flora Salim’s invited talk titled “Towards World Models for Urban Mobility” at a workshop on Urban Mobility Foundation Models.
“My biggest takeaway is that SIGSPATIAL showcases a diverse range of interconnected research, and it was encouraging to see that my PhD research questions remain both open and highly relevant challenges” Wilson said.
“I also gained new ideas on methods and approaches that I can potentially apply to my research.”
Wilson then visited the HuMNet Lab at the University of California, Berkeley. He was hosted by Professor Marta C. Gonzalez, whose research focuses on urban mobility. At Berkeley, Wilson presented his PhD work so far, including GenUP and Massive-STEPS to Professor Gonzalez and her team.
The presentation sparked in-depth discussions with her students about their research, spanning topics such as clustering human lifestyles from mobility traces, examining geographical biases in existing systems, and applying classical urban mobility theories to modern LLM approaches.
As a result, Wilson confirmed an upcoming collaboration with one of Professor Gonzalez’ students. He plans to contribute modern machine learning techniques alongside classical theoretical approaches.
“I aim to dive deeper into classical urban mobility theories, as my background in cutting-edge LLMs can overlook these foundational concepts, combining modern models with classical theories will allow us to build more robust and explainable ‘hybrid’ systems.” Wilson said.
“It was inspiring to see that we are tackling the same research problems in parallel, each leveraging our own strengths and perspectives, collaborating in this way yields meaningful and impactful results.”
Wilson’s research on this trip is part of the broader GenAISim Signature project at ADM+S. This trip was funded through ADM+S HDR funding and the GenAISim project.
