Professor Flora Salim, Cisco Research Chair in Digital Transport and AI, of the School of Computer Science and Engineering has won a competitive Cisco Research Gift. The Gift will be used to develop a new Mobility Question Answering engine for general spatio-temporal forecasting tasks.

The aim is to use mobility data to answer questions like: “How crowded will Central Station be this Saturday after the concert?”, “How many customers will there be in the shopping mall next Monday morning?” and “Where can I likely get a parking spot near the cafe tomorrow lunchtime?”.

Question-answering applications are currently found in various services, such as navigation chatbots. However, this is the first project of its kind that will use the Q&A setting to answer forecasting questions.

The project will use language models to describe heterogeneous spatio-temporal and mobility data in a natural language. If deployed, it will enable robust spatio-temporal forecasting, empowering both transport operators and end users in their daily decision-making around transport and mobility.

“Winning this gift shows the significance of this research problem, as well as the strong interest and novelty of the research idea,” said Prof. Salim.

“We are happy to learn that Cisco Research, US, is interested in research into enabling a more natural and effortless interaction with big IoT data, as well as forecasting tasks through a question-answering pipeline,” she said.

Jayanth Srinivasa, Senior Cisco Researcher, said "We are excited to support Prof. Salim's pioneering work using language models in spatio-temporal forecasting. We believe this research has the potential to address real-world challenges and improve the way people interact with mobility data."

Prof. Salim’s research lies in the intersection of human-centred computing and machine learning for behaviour intelligence, towards scalable and generalizable behavioural AI on the edge. Applications include intelligent assistants, health monitoring, smart cities, smart buildings, and spatio-temporal recommendation systems.

The initial works that led to the funding of this gift were co-authored with Dr. Hao Xue, and published at the Proceedings of the Fifteenth ACM International Conference on Web Search and Data Mining (ACM WSDM 2022)1 and the Proceedings of the 30th International Conference on Advances in Geographic Information Systems (ACM SIGSPATIAL 2022)2.

Professor Salim won the Women in AI Awards 2022 ANZ in the Defence and Intelligence category. She is a member of the Australian Research Council (ARC) College of Experts, a Chief Investigator of ARC Centre of Excellence for Automated Decision Making and Society, an Editor of Proceedings of the ACM on Interactive, Mobile, Wearable, Ubiquitous Technologies (IMWUT), the Associate Editor-in-Chief of IEEE Pervasive Computing, and an Associate Editor of ACM Transactions on Spatial Algorithms and Systems.

She was a Humboldt-Bayer Fellow, Humboldt Fellow, Victoria Fellow, and ARC APDI Fellow. She obtained her PhD from Monash University in 2009.

[1] Translating Human Mobility Forecasting through Natural Language Generation

[2] Leveraging language foundation models for human mobility forecasting