From Risk Recognition to Recovery: A Human-Centered Approach to Supporting Digital Financial Safety Among High-Risk Populations

The Hong Kong University of Science and Technology
Department of Computer Science and Engineering


PhD Thesis Defence


Title: "From Risk Recognition to Recovery: A Human-Centered Approach to
Supporting Digital Financial Safety Among High-Risk Populations"

By

Miss Yue DENG


Abstract:

Digital financial safety is often treated as a technical problem of 
preventing unauthorized transactions, detecting warning signs, and reducing 
financial loss. Yet financial risk is also socio-technical, shaped by 
people's interactions with digital systems and others. Taking a human- 
centered approach, this dissertation examines how people perceive digital 
financial risks, how they receive social and technological support when 
facing such risks, and how they recover when harm cannot be prevented, 
particularly on high-risk populations, and builds intelligent systems to 
support these processes. First, I investigate older adults' risk perceptions 
in emerging financial services. Through their experiences with biometric 
payment, I show that frictionless interactions can obscure authentication and 
authorization, making it difficult to understand what is happening, what may 
go wrong, and what control remains. This creates a gap between technical 
protection and perceived safety. I then examine how high-risk populations can 
be supported when facing digital financial risks such as online fraud. 
Through studying family safeguarding, I find family members often act as an 
informal security infrastructure by monitoring, advising, detecting scams, 
and intervening. However, this support is constrained by refusal of help, 
delayed disclosure, interpersonal tension, and emotional and financial 
burdens. Digital financial safety therefore cannot depend entirely on the 
continued involvement of family members. Building on this insight, I develop 
and evaluate a role-based anti-fraud intervention that helps older adults 
learn from multiple scam perspectives and strengthen their ability to 
recognize and respond to fraud. To address what happens after victimization, 
I develop ReGround, an AI-mediated system that extends post-scam support 
beyond financial containment to emotional recovery, helping victims cope with 
psychological harm and return to everyday life. Together, these contributions 
show that protecting high-risk populations requires moving beyond technical 
safeguards. By integrating risk perception, social support, capability 
building, and post-harm recovery, this dissertation presents digital 
financial safety as an ongoing process from risk recognition to recovery.


Date:                   Tuesday, 28 July 2026

Time:                   10:00am - 12:00noon

Venue:                  Room 3494
                        Lifts 25/26

Chairman:               

Committee Members:      Prof. Bo LI (Supervisor)
                        Dr. May FUNG
                        Dr. Xiaojuan MA
                        Dr. Ningning DING (HKUST-GZ)
                        Prof. Jianliang XU (HKBU)