Mitigating Urban Crises via the Lens of Human Mobility Data: Measurement and Intervention

PhD Thesis Proposal Defence


Title: "Mitigating Urban Crises via the Lens of Human Mobility Data: 
Measurement and Intervention"

by

Miss Lin CHEN


Abstract:

Cities are complex sociophysical systems marked by heterogeneity, 
interconnectivity, and dynamicity. These features pose unique challenges in 
tackling urban crises such as pandemics, natural disasters, environmental 
degradation, and infrastructure inadequacy. Addressing such crises requires a 
comprehensive understanding of what people experience and how they react in 
complex urban environments. Human mobility data, collected from widely-adopted 
GPS-equipped mobile devices, provide a powerful lens of study with its 
large-scale, fine-grained, and real-time nature. In this thesis proposal, I 
develop data-driven methodologies to harness human mobility data for sensing, 
understanding, and intervening in urban crises.

The plight of "left behind" populations in urban crises is a pressing concern, 
but a comprehensive picture of their situations is lacking, hindering effective 
intervention. In the first work, we construct VulnerabilityMap, an open data 
framework to sense the impact of various urban crises on disadvantaged 
populations. We propose a taxonomy to categorize urban crises into a spectrum 
of shocks and stresses inspired by Maslow's hierarchy of needs. Under the 
taxonomy, we align and refine human mobility data with other demographic, 
neighborhood, and online behavior data, to reflect people's potential 
disadvantages and actual vulnerability outcomes to crises. With our framework, 
we train machine learning models that exhibit strong prediction performances 
and offer insights into the social factors trapping certain populations in 
vulnerable situations.

Emergent crises typically go through a life cycle of surge and diminishment. In 
the second work, we conduct a thorough data-driven analysis to understand urban 
social dynamics across different phases of the COVID-19 pandemic at scale. We 
uncover two typical patterns that govern mobility changes in most urban venues: 
daily life-centered venues experience smaller reductions and quicker recovery, 
while work-centered venues face larger reductions that persist. These changes 
exert deeper impacts on the underlying social fabric, significantly correlating 
with income and racial segregation. Through clustering analysis, we further 
uncover the heterogeneity in mobility changes across geographical regions that 
are associated with partisan inclinations.

Fair and effective resource distribution is crucial in crisis intervention. In 
the third work, we study this task with the case of COVID-19, focusing on 
data-driven epidemic modeling and intervention simulation. We propose a 
mobility data-driven epidemic model that explicitly captures both demographic 
and behavioral differences among communities to explain heterogeneous risks. By 
simulating different distribution strategies, we find that prioritizing 
vaccines for the most disadvantaged communities can simultaneously improve 
social utility and equity, even with considerable vaccine reluctance. 
Nevertheless, equity across demographic dimensions may conflict. To solve this, 
we propose an automatic framework that adaptively distributes vaccines to 
balance multiple ethical values.

In the fourth and ongoing work, we take a further step to dissect the role of 
mobility network structure in explaining urban social resilience during crises. 
By constructing POI sector networks based on co-visitation patterns, we 
discover a strong and positive correlation between network centrality and 
segregation. Such patterns are consistent across multiple types of crises.

We conclude this thesis proposal with future research directions and challenges 
related to data-driven urban crisis management.


Date:                   Wednesday, 25 September 2024

Time:                   5:00pm - 7:00pm

Venue:                  Room 5501
                        Lifts 25/26

Committee Members:      Prof. Pan Hui (Supervisor)
                        Dr. Tristan Braud (Co-supervisor)
                        Prof. Gary Chan (Chairperson)
                        Prof. James Kwok
                        Dr. Gareth Tyson (HKUST-GZ)