UWSpace
UWSpace is the University of Waterloo’s institutional repository for the free, secure, and long-term home of research produced by faculty, students, and staff.
Depositing Theses/Dissertations or Research to UWSpace
Are you a Graduate Student depositing your thesis to UWSpace? See our Thesis Deposit Help and UWSpace Thesis FAQ pages to learn more.
Are you a Faculty or Staff member depositing research to UWSpace? See our Waterloo Research Deposit Help and Self-Archiving pages to learn more.

Communities in UWSpace
Select a community to browse its collections.
- The University of Waterloo institution-wide UWSpace community.
Recent Submissions
Item type: Item , Distilling Legal Language Models with Limited Data through Synthesis and Attribution(University of Waterloo, 2026-11-20) Arbabi, AlirezaLarge language models are increasingly capable on specialized legal NLP tasks, but their direct use incurs recurring inference costs and may require proprietary or confidential data to be shared with external providers, creating practical barriers for sensitive legal applications. In this setting, knowledge distillation provides a natural approach for building small and specialized language models. However, conventional distillation methods are limited when only a sparse set of real task examples is available, a common challenge in the legal domain. To address this challenge, this thesis presents LegalSynth, a framework that uses synthetic data generation followed by attribution-based data selection as an alternative channel for transferring knowledge from the teacher to the student. Given a task description and a small set of real examples, LegalSynth first prompts a teacher model to infer the task structure from the available task information and then leverages this understanding to generate a large and diverse pool of synthetic training examples beyond the initial seed set. Because synthetic examples vary in quality and utility, LegalSynth then applies data attribution to filter the synthetic pool and retain the examples that are most useful for training, improving the performance of the student model. Experiments across four low-resource, multi-way legal classification tasks from LegalBench demonstrate the effectiveness of this framework, enabling Qwen3.5 models with up to 4B parameters to compete with, and in several settings outperform, frontier language models including GPT-5.6-Terra, Gemini-3.7-Flash, and Qwen3.5-397B.Item type: Item , Examining the takeover process for senior drivers in conditionally automated vehicles(TS International Journal, 2024) Ding, Wen; Murzello, Yovela; Samuel, Siby; Cao, ShiWhile the consensus suggests that autonomous control is particularly advantageous for older drivers due to reduced driving workload, the challenges brought on by the takeover process may impose additional burdens on this cohort. This study aims to explore the takeover time and takeover quality of older drivers, examining how different driving conditions affect their performance compared to young and middle-aged drivers. By doing so, it seeks to pinpoint potential risks specific to this demographic segment. Seventy-two participants from three age groups - younger, middle-aged, and older drivers - were recruited, each compelting eight takeover scenarios. This study addresses a gap in age-related takeover studies, which predominantly target young-old age drivers (65-74 years old), by focusing specifically on the old-old age group (75+ years old) characterized by heightened cognitive impairments. The result revealed that older drivers have significanrly higher takeover time and lower takeover quality compared to younger and middle-aged groups. Older drivers exhibited greater use of acceleration and brake more frequently after takeover, and they had significantly higher lateral drifting. Furthermore, the non-driving related tasks distract older drivers more copared to drivers from other age groups. Highway and local roads, as well as curved and straight roads, affect drivers across all age groups similarly. The findings suggest that older drivers encounter enhanced challenges during the takeover process, particularly when non-driving-related tasks were presented during the automation control. Although automation provides convenience to older people, attention should be directed directed towards mitigating the additional risks associated with takeovers. This conclusion contributes valuable insights for the future customized design of takeover requests tailored to older drivers.Item type: Item , Psychophysiological response, user experience and performance in routine and emergency flight simulation training scenarios using multimodal wearable sensing(Springer, 2026) Shafi, Anika Fairooz; Munoz, John Edison; Cao, ShiWearing sensors are continually extending the possibilities for physiological monitoring and physio-adaptive (biocybernetic) systems, with particularly relevant applications in pilot training. This study exporing psychophysiological responses, performance and user experience of pilot trainees (n = 16) as they interacted with routing and high-stress flight simulation scenarios. Outcome measures include electrodermal activity (EDA), heart rate (HR) and heart rate variability (HRV) parameters, competency-based performance metrics and perceived user experience and stress. Sympathetic activation was observed, characterized by elevated HR and reduced HRV time-domain measures (Mean NN, RMSSD, and pNN50) alongside decreases in non-linear measures (SD1 and the SD1/SD2 ratio) across progressively more stressful conditions. Self-reported stress significantly differed across conditions, while EDA data showed notable individual variability. Performance, measured by mean application of procedure scores, was significantly affected. Results indicate initial evidence that the multimodal wearable physiological sensors employed in the study sensitively and consistently gauge changes in stress inducements. These findings have important implications for stress inoculation training, personalized instructor feedback and development of future biocybernetic training systems.Item type: Item , Current practice and challenges in coaching Esports players: An interview study with league of legends professional team coaches(Elsevier, 2022) Sabtan, Bader; Cao, Shi; Paul, NaomiEsports is rapidly growing around the globe. The need for coaching best practice is becoming essential with growing investments in teams and players. In this paper, we investigated the current coaching practices and challenges in League of Legends professional teams by interviewing professional coaches from international teams. Our results showed unanimous agreement that coaching in Esports remains in its early stages, with a need to develop standards for coaching. We were able to understand the role of coaches, their different positions, and their skill requirements. We were also able to learn the challenges players face such as long practice hours, player attitudes, mental stress, and other health concerns such as back and wrist injuries. From the interview, we identified nine aspects of research requiring future studies.Item type: Item , Exploring perceptions of pilot licensing and training standards: A survey of Canadian student and licensed pilots(Taylor & Francis, 2025) Rizvi, Syed A. Q.; Moncion, Bradley; Cao, ShiThis study investigates the perspectives of Canadian student and licensed pilots on general aviation pilot training and licensing practices. Employing a mixed-methods approach, this research critiques the reliance on flight hours as the sole competence metric and examines the alignment of existing practices with modern aviation's complexities. Findings reveal a divergence in opinions between novice and experienced pilots on flight hours' importance, with a consensur towards a competency-based evaluation model. The study identifies critical shortcomings in existing training practices, such as the challenge of integrating technology, fostering advanced skills, and efficiently utilizing instructional resources. It suggests recommendations for regulatory enhancements, aiming to ensure training practices evolve in line with the changing requirements for aviation safety and technology. The conclusion calls for urgent reform, underlining the imperative for training adaptations that can prepare pilots to proficiently manage the complexities of contemporary airspace, thus safeguarding their proficiency and safety.