Memorial University Research Repository

The Memorial University Research Repository is an open access initiative to showcase and preserve Memorial University's creative and intellectual output, including theses, journal articles, conference papers, lectures, presentations, reports, and performances.

Communities in Memorial University Research Repository

Select a community to browse its collections.

Now showing 1 - 18 of 18

Faculty of Business Administration 12
Faculty of Education 97
Faculty of Engineering and Applied Science 311
Faculty of Humanities and Social Sciences 259
Faculty of Medicine 314
Faculty of Nursing 14
Faculty of Science 835
Grenfell Campus 322
Labrador Campus 4
Marine Institute 143
Memorial Theses 11524
Memorial University Affiliates 95
Memorial University Libraries 145
School of Human Kinetics and Recreation 11
School of Music 14
School of Pharmacy 16
School of Social Work 8
Specialized Centres 334

Recent Submissions

Access status: Open Access ,
Dissociating the self: representations of mental illness in graphic memoir
(Memorial University of Newfoundland, 2026-02) Velentzas, Irene; Pedri, Nancy
Although the field of graphic medicine is burgeoning, the rapid proliferation of mental illness graphic memoir over the past two decades has yet to receive sustained scholarly attention (Velentzas 2021). This dissertation aims to correct that oversight by examining how mental illness graphic memoir operates as an essential vehicle for engaging with verbal-visual representations of mental illness through pictorial embodiment, graphic style, verbal-visual tensions, factors of visual coherence, visual metaphor, multiple textual delivery systems, and self-reflexivity. This study relies on foundational scholarship in the fields of disability studies, life-writing, comics formalism, and graphic medicine to examine the underlying constructions of mental illness representations in graphic memoir through close reading analysis and “research creation” (Loveless 2019). The dissertation’s academic portion closely examines seven mental illness graphic memoirs – Becoming Unbecoming (2016), Depresso (2010), Hyperbole and a Half (2013), Inside Out (2007), Lighter than My Shadow (2013), Marbles (2012) Solutions and Other Problems (2020) – alongside seventeen additional texts to determine patterns of representation that address and challenge stigmatic mental illness language and constructions. Its creative component, the graphic memoir Undiagnosed (2025), further applies and extends the foundational theory and the dissertation’s academic findings. “Dissociating the Self: Visual and Verbal Representations of Mental Illness in Graphic Memoir” reveals the ubiquitous use of a representational strategy I term dissociation – a visual mode that communicates the separation of the narratorprotagonist’s understanding of self from stigmatic understandings of mental illness embodied by a double. Visual dissociation is shown to operate through different stylistic appearances of the self and the double (Chapter 1), which encodes the moral, medical, and social paradigms of disability (Chapter 2). The double dissociates from stigmatic understandings of mental illness through the use and conceptual alteration of common visual metaphors for encoding mental illness, such as monsters and darkness (Chapter 3 and 4). Following the introduction (Chapter 5) and examination of Undiagnosed (Chapter 6), several verbal dissociation strategies are noted operating in mental illness graphic memoir, including differentiated font styles for personal and medical narratives; purposeful silence; disembodied speech balloons; and the inclusion, ironizing, and overwriting of medical texts (Chapter 7). Ultimately, “Dissociating the Self: Visual and Verbal Representations of Mental Illness in Graphic Memoir” demonstrates and theorizes the unique multimodal affordances of mental illness graphic memoir that enable cartoonists to encode, challenge, and overturn stigmatic understandings of mental illness to construct new understandings founded on empathy.
Access status: Open Access ,
Agricultural management in boreal regions alters soil respiration burst profiles
(Memorial University of Newfoundland, 2026-02) Vallotton, Jeremiah Daniel; Unc, Adrian
Land use conversion and climate change represent major threats to carbon (C) storage in boreal soils. Chapter I (literature review) of this dissertation shows that natural boreal soils are well-equipped to handle climate change, given that these ecosystems are already shaped by disturbance and possess natural mechanisms to resist or compensate for C loss under climate change. Much less is known about converted boreal soils under agricultural management, despite rapid and ongoing conversion. Chapters II-V of this dissertation thus combined four studies at different resolutions (within-field, within-farm, within-region, and global) to examine the fate of soil C in converted boreal soils through intensive soil abiotic property and soil respiration testing. Chapter II (within-field) examined a chronosequence of converted organic and podzolic soils and found that agricultural management rapidly homogenized soil to agricultural norms but also lost C rapidly under mineral management; surprisingly, high C presence resisted homogenization. Chapter III (within-farm) compared paired agricultural fields under normal management or amended with pulverized rock, and found that pulverization does not affect soil properties immediately; soil pedogenic type was more important in determining behavior. Chapter IV (within-region) compared soils under four land uses (LUs), and found that respiration patterns in each LU were distinct enough from each other (forest, agriculture, grassland) to allow for accurate prediction of LU. Chapter V (global) compared four LUs worldwide (agricultural, grassland, forest, and transitional) and found that global patterns of forest and agriculture could be accurately predicted via respiration, but grassland and transitional were too heterogeneous. Furthermore, respiration could indicate underlying management effects more accurately than abiotic factors on the global scale. This dissertation reveals that 1) boreal land conversion can cause <5yr shifts to agricultural norms accompanied by rapid C loss; 2) traditional management of boreal abiotic fertility fails to address questions of soil health and C cycling; 3) the status of boreal soils can be understood through LU-specific patterns of proportional respiration; and 4) conversion to agriculture creates anthropic soils with distinct behavior patterns that can be accurately predicted globally and must thus be managed differently from natural sites.
Access status: Open Access ,
A pilot study of a novel sleep scoring system to measure insomnia treatment response in breast cancer survivors
(Memorial University of Newfoundland, 2026-02) White, Emily A.; Garland, Sheila
Insomnia is prevalent among breast cancer survivors and Cognitive Behavioural Therapy for Insomnia (CBT-I) is an effective treatment. Research using polysomnography (PSG) to objectively measure sleep outcomes in response to CBT-I is limited. This single-arm study of nine breast cancer survivors examined sleep response to CBT-I using an in-home PSG device. The first objective examined feasibility of using the Cerebra Sleep System, an in-home PSG device, pre- and post-treatment. Recruitment and retention rates were relatively low, but the device was feasible to use. Attitudes towards using the device were mixed; some felt it was fine while others felt it was awkward and may have impacted their sleep. The second objective examined CBT-I sleep outcomes measured with sleep diaries and in-home PSG. Sleep diary measures of sleep onset latency, sleep efficiency, and wake after sleep onset significantly improved. PSG-measured sleep efficiency and wake after sleep onset significantly improved but time spent in sleep stages did not significantly change. Using an in-home PSG device may be feasible with changes to improve recruitment and retention rates and lessen the burden on participants. CBT-I may result in objective improvements in sleep continuity metrics. Future research should consider a largescale study with changes in methodology.
Access status: Open Access ,
Social, lifestyle, and genetic determinants of multimorbidity clusters in middle-aged and older Canadian adults: an analysis of the clsa data
(Memorial University of Newfoundland, 2026-02) Mortey, Obed; Gao, Zhiwei; Mugford, Gerald; Aubrey-Bassler, Kris; Mariathas, Hensley H.
Background: Multimorbidity, defined as the co-occurrence of two or more chronic conditions, is increasingly prevalent among aging populations. Although lifestyle and socioeconomic factors contribute substantially, the genetic underpinnings of multimorbidity remain poorly understood. Objectives: This thesis aimed to identify the most common multimorbidity (MCM) among middle-aged and older Canadian adults, and to investigate the prevalence, associated risk factors, and genetic susceptibility of MCM. Methods: This study included 30,097 participants from the comprehensive cohort of the Canadian Longitudinal Study on Aging. Survey-specific multivariate logistic regression was used to identify significant risk factors of MCM. A polygenic risk score (PRS) was derived for each participant. The modification effects of the PRS on the association between age and risk of MCM were examined by a Genome-wide interaction study. Results: Osteoarthritis–hypertension emerged as the MCM with a prevalence of 16.5% among middle-aged and older Canadian adults. Seven factors including increasing age, retirement, poor perceived health, sleep problems, obesity, urban core residence, and living in eastern provinces were significantly associated with increased risk of MCM. Ten genetic variants with an interaction term p-value <10-5 were selected to be included in the calculation of PRS for each participant. The participants in the top PRS tercile (top 1/3 PRS) exhibited the greatest risk of MCM. For each additional year of age, MCM risk increased by 13% (Adjusted Odds Ratio (AOR)=1.13, 95% Confidence Interval (CI): 1.11 – 1.15) in the top-PRS group compared with 9% (AOR=1.09, 95%CI: 1.07 – 1.12) in the middle PRS tercile and 10% (AOR=1.10, 95%CI: 1.08 – 1.12) in the low PRS tercile group. Conclusion: Both social-environmental and genetic determinants jointly influence multimorbidity, highlighting the need for integrated prevention strategies and precision aging interventions in Canada.
Access status: Open Access ,
Machine learning for malware and intrusion detection: dataset design, cost-aware models, and research pitfalls
(Memorial University of Newfoundland, 2026-02) Kamyabi, Javad; Anderson, Jonathan
Information technology has reduced constraints of physical distance and delays associated with traditional methods in areas such as medicine, economy, industry, and beyond. However, it also presents potential threats such as hackers and cybercriminals. As information technology advances, threats become smarter and more complex, cat-and-mouse-game that continuously increases in complexity. Machine learning improves security tools such as malware or intrusion detection by taking advantage of past experiences. Machine learning requires high-quality datasets to create effective models. The first paper in this thesis, eBPF-Powered Dynamic Analysis for Linux Malware Detection: A Dataset and Experimental Study, explores the application of machine learning to detect malware. The paper also introduces an automated eBPF-based data collection pipeline using Docker containers to generate labeled malware and clean environment traces. We construct a dataset of clean and infected Linux operating systems and use various machine learning techniques to identify patterns in Linux system calls that indicate whether the operating system is infected or not, achieving a detection F1-Score of up to 99% with Random Forest models. Machine learning can also be used to develop intrusion detection systems. Two critical components of such systems are the dataset and the models. However, popular network attack datasets suffer from imbalances, with significant disparities in the number of instances between different classes (e.g., benign traffic can have thousands of samples, while rare attack types may have fewer than 50). This imbalance can severely affect model performance; for example, rare attack classes may be underrepresented by a ratio of 40:1 compared to benign traffic, which can significantly reduce recall for these classes. To address this issue, over- and undersampling methods balance datasets before feeding them into the algorithms. However, undersampling may overlook important data, whereas oversampling can introduce redundancy, ultimately weakening the model's performance. Furthermore, the speed with which an intrusion detection tool makes decisions plays a vital role in its effectiveness. The second paper in this thesis, titled Cost-Aware Machine Learning for Intrusion Detection: A Performance Trade-Off Study, demonstrates that by sacrificing an insignificant amount of accuracy, it is possible to achieve models that are tens of times faster and significantly less memory-consuming, making them practical for real-time deployment. This is accomplished by exploring the combination of different deep learning and machine learning models, along with various over- and under-sampling methods. Furthermore, the paper proposes twelve prediction cost functions that integrate these trade-offs alongside traditional performance measures. A slow intrusion detection tool can otherwise become a bottleneck in a network, highlighting the need for models that balance accuracy and efficiency. The third paper, titled Power and Pitfalls of ML-Based Intrusion Detection Systems, examines key challenges in developing machine learning-based intrusion detection systems, with a focus on both dataset generation and model design. It highlights issues such as the lack of representative datasets and the limited generalizability of models. This paper examines ten significant research barriers and their interconnections, which means that a barrier may lead to one or more barriers. The study includes a statistical analysis of dozens of research papers, revealing the current state of the field. Two best-practice checklists are proposed to guide future work in dataset creation and IDS research, with the aim of improving the quality and reliability of publications in this domain. Together, these three studies provide a comprehensive framework for designing more accurate, efficient, and trustworthy machine learning-based security tools such as malware or intrusion detection systems. By combining practical data generation, cost-aware modeling, and critical analysis of research pitfalls, this thesis contributes to more robust and realistic security research and practice.