39339 research outputs found
Sort by
From Google to Generative AI: Rethinking the role of library research guides
No full textSince the introduction of ChatGPT in 2022, many studies have reported decreases in traffic to websites such as Google and Wikipedia, reflecting broader shifts in user information-seeking behaviour. To understand how these changes may affect academic libraries, we analyzed LibGuides traffic statistics and data from Google Analytics at our institution from their launch in 2015 through the present. Our study compares usage patterns before and after the emergence of ChatGPT, identifies where students accessed the guides from, and pinpoints which guides continue to be actively used in the era of generative AI. This investigation aims not only to trace the impact of tools like ChatGPT on library research guides but also to generate insights into how guides should be tailored and redesigned to support students’ evolving mental models and information-seeking behaviours.
This presentation was delivered at 2026 OLA Super Conference in Toronto on January 30, 2026
Towards Enhanced Resiliency and Interoperability of LoWPAN-Based Smart Grid Applications
No full textEffective electrification is a cornerstone of modern living and its significance is expected to intensify with the ongoing urbanization and densification of residential areas. In particular, the electrification of the residential sector has emerged as a critical challenge, as recent North American construction trends show a steady shift toward high density, multi-unit developments and a decline in single family housing permits. This transition amplifies the demand for scalable, reliable and cost efficient grid infrastructure capable of supporting dense clusters of electrical loads and distributed energy resources. Achieving such transformation requires robust communication infrastructures, which form the backbone of essential services including power grid load balancing, smart meter data collection, solar energy management and Electric Vehicle (EV) charging. Ensuring the affordability, reliability and performance of these interconnected subsystems is therefore fundamental to realizing the full potential of multi-residential electrification.
Although traditional communication technologies such as Wi-Fi are mature and widely available, their deployment across relatively large or enterprise scale networks including multi-residential buildings, often proves cost prohibitive. Furthermore, Distributed Energy Resource (DER)s, smart meters and Electric Vehicle Supply Equipment (EVSE)s do not necessarily demand the high bandwidth provided by such technologies and can instead benefit from alternative paradigms that offer adequate performance at a lower cost. Narrowband communication technologies emerge as a promising alternative, presenting inherent challenges due to their lossy nature, which necessitates the implementation of efficient error correction mechanisms to maintain reliable operation. In a continuously data driven world, selecting and tuning the appropriate error correction strategy is paramount. Given the resource constrained nature of embedded hardware, these mechanisms must also remain cost effective and computationally efficient to ensure practical deployment.
When properly configured, deployed and supported by robust error correction mechanisms, narrowband communication demonstrates substantial potential to enhance the scalability, reliability and overall practicality of electrification initiatives in multi-residential buildings. This thesis presents a holistic study on the scope of narrowband network resiliency and interoperability, addressing critical gaps in existing Low-Power Wireless Personal Area Network (LoWPAN) implementations regarding robust connectivity and data imputation. By treating network architecture and communication aware data continuity as coupled requirements rather than independent topics, the work advances the current body of knowledge and ensures practical viability. Consequently, it offers an end-to-end evaluation framework tailored for the complex realities of multi-residential smart grid operations
Trapping Efficiency of Non-Buoyant Microplastics by River Groynes
No full textMicroplastics (MPs) have become an urgent environmental issue, driven by rapid industrial plastic use and inadequate waste management. As rivers act both as major transport pathways and temporary sinks, reducing MPs loads within them can effectively limit their transfer into the food chain. This study investigates the effectiveness of groynes, structures originally designed for flow and erosion control, in trapping non-buoyant MPs. Ten configurations formed from pairwise combinations of four common groyne geometries were tested using MPs (of equivalent diameter d_eq = 2.29 mm) of two densities (ρ_s = 1.08 and 1.11 g/cm³). Experiments were conducted in a recirculating flume using subcritical turbulent flow. A particle tracking model was employed to track and record particle trajectories around groyne fields, from which trapping efficiencies and retention zones were derived. It was observed that MPs entered the groyne field from the downstream end, forming a clockwise gyre. All scenarios retained > 6% of particles during the test period. When both groynes used in the experiments shared the same shape, upstream-facing inclined groynes exhibited the greatest trapping potential, achieving a peak retention rate of 20.7 ± 3.7% over extended durations. Despite their lower entrainment, straight and T-shaped groynes retained particles more effectively, with < 3% trapped particles leaving the field during the tested period. In cases where the upstream groyne was straight and the downstream geometry was varied, particle escape was negligible, with downstream repelling groynes consistently trapping more MPs than other configurations. Although particle density had little effect on particle entrainment, lighter particles showed lower escape and smaller retention zone, indicating MPs sensitivity to vertical flow dynamics within the groyne field
Using Learning Analytics & Machine Learning to Enhance Early Detection of At-Risk Students in Higher Educational Institutions
No full textThe rapid transition to remote learning during the Coronavirus Disease of 2019 (COVID-19) pandemic significantly disrupted student-instructor interaction, contributing to increased dropout rates in higher education. According to the 2022 Canadian Student Wellbeing Study, 72% of students experienced reduced in-person engagement, and 40% considered withdrawing due to insufficient institutional support. In response, Learning Analytics (LA) has gained prominence as a data-driven approach to enhancing student engagement, academic performance, and retention. This thesis addresses two core challenges: the need for a scalable LA framework and the development of an effective Machine Learning (ML)-based LA to identify at-risk students.
Despite growing interest in LA, many studies overlook foundational implementation challenges and the limited capabilities of existing LA tools, such as Brightspace’s Student Success System (S3). To address these gaps, this study begins with a review of LA adoption in global and Canadian contexts, highlighting that while over half of Canadian Higher Education Institutions (HEI) are engaging in LA initiatives, empirical evidence of impact remains scarce. The thesis proposes a comprehensive LA architecture tailored to The University of Ottwawa (uOttawa), emphasizing early risk detection and addressing functional gaps in current Learning Management System (LMS) tools.
Building on this foundation, the core of this work involves the development of ML model for at-risk student prediction. Following the Cross-Industry Standard Process for Data Mining (CRISP-DM), six classification algorithms are evaluated across four temporal checkpoints in the academic term. The models incorporate features from the Student Information System (SIS) and Brightspace LMS, including demographic attributes, academic history, and time-aware performance metrics. Feature importance analysis reveals a dynamic shift from reliance on historical SIS data to LMS-derived grade indicators as the course progresses. Notably, time-based features—especially phase-specific grade averages—proved crucial for late-phase predictions. Among the models tested, Random Forest (RF) and Extreme Gradient Boosting (XGB) consistently achieved the highest recall and accuracy, identifying 75–91% of at-risk students with overall accuracy between 60–81%.
This thesis offers three key contributions: (1) a critical review of LA adoption and implementation strategies; (2) a proposed institutional LA architecture to support predictive analytics; (3) an empirical methodology of comparing classification models and time-aware feature dynamics, as a foundation for a deployable Early Warning System (EWS) to inform proactive interventions for improving student success and retention
Nuclear-Trafficked Amphiregulin Regulates H3K9me3 Heterochromatin Formation Facilitating Replication Control and Mitigating Replication Stress-Associated DNA Damage in Mammary Epithelial Cells
No full textThe epidermal growth factor receptor (EGFR) ligand, amphiregulin (AREG) is widely expressed in epithelial cells. It is synthesized as a precursor protein that undergoes proteolytic shedding from the plasma membrane. In mammary epithelial cells (MECs), AREG engages in paracrine signaling where it functions as an important differentiation factor during mammary development at puberty. Interestingly, AREG ectodomain shedding also stimulates a lesser known, nuclear trafficking pathway, wherein it bypasses the lysosome and interacts with lamin A/C at the inner nuclear membrane (INM) and promotes H3K9me3 heterochromatin formation. However, the physiological stimulus, role, and molecular mechanism underlying AREG-mediated H3K9me3 maintenance have not been elucidated. Here, we show that in response to replication stress (RS), AREG undergoes retrograde trafficking to the nucleus in a p38-dependent manner via early endosomes in hTERT-immortalized MECs. We found that endocytosed AREG interacts with prelamin A in the cytoplasm and nucleus correlated with a transient increase in H3K9me3. siRNA-mediated AREG knockdown (KD) resulted in the downregulation of both suppressor of variegation 3-9 homolog 1 (SUV39H1), and heterochromatin protein 1 alpha (HP1α) accompanied by the reduction and decompaction of H3K9me3. Notably, AREG depletion resulted in increased DNA damage in BRCA2+/+ hTERT-MECs that was exacerbated in BRCA2mut/+ hTERT-MECs, following the inducing of RS. AREG-KD in these MECs increased DNA lesions associated with perturbed replication dynamics including slowing of the replication fork and enhanced firing of new origins. Depletion of AREG also disrupted the Ran-gradient, weakened nuclear membrane integrity resulting in the appearance of cytoplasmic DNA and activation of interferon (IFN) pathway associated genes, ultimately leading to senescence. Conversely, ectopic expression of AREG-ΔC11 lacking the C-terminal 11 amino acids, which traffics exclusively to the nucleus, increased prelamin A levels at the nuclear periphery coinciding with an increase in SUV39H1 and HP1α. Together, these experiments demonstrated a non-canonical role for AREG in maintaining heterochromatin, possibly by facilitating nuclear import of prelamin A. This process is strongly enhanced by RS, resulting in accumulation of unprocessed nuclear prelamin A and a transient increase in heterochromatin. Integrating our findings and the known ovarian hormone-dependent regulation of AREG and its role in promoting mammary progenitor differentiation, we propose that AREG safeguards genomic stability and lineage fidelity during reproductive years. After menopause, a decline in extracellular AREG may contribute to the loss of lineage fidelity and reduced heterochromatin maintenance in some MEC subpopulations. This reduction in chromatin integrity promotes DNA damage in proliferative MECs and activates IFN pathway leading to senescence in both mature and cycling MECs. Collectively, these alterations create a pro-tumorigenic environment that may contribute to the marked increase in incidences of breast cancer in postmenopausal women, a population that accounts for 70% of all breast cancer cases
Creating an integrated innovation system to enable the adaptation and uptake of health-system innovations in Canada: insights from citizen panels and a national stakeholder dialogue
No full textAbstract Background Health-system leaders are increasingly faced with making decisions about whether and how to use a wide range of current and emerging health-system innovations to address complex system and policy challenges. Health-system innovations can broadly include new ways of doing things at a system level, such as new approaches to govern health systems, care delivery, funding models, health policy or better ways to integrate health and social services. However, Canada has historically struggled with the adaptation and uptake of health-system innovations. This multicomponent study aimed to explore the challenges, approaches and implementation considerations for creating an integrated innovation system that enables the adaptation and uptake of health-system innovations from the perspectives of citizens and health system leaders in Canada. Methods We synthesized the best-available evidence into an evidence brief and a subsequent plain-language version (a citizen brief) in consultation with a steering committee and key informants, including policymakers, leaders of systems, organizations and professional organizations, industry representatives, citizen leaders and researchers. These briefs informed deliberations in four citizen panels (n = 48 participants) and a national stakeholder dialogue with health-system leaders (n = 23 participants) to identify key challenges, approaches, implementation considerations and next steps that could be taken. Results Citizen panel participants and health-system leaders highlighted barriers such as culture and mindsets that resist health-system innovations, limited targeted funding for health-system innovations and processes that encourage sustainability, lack of mechanisms to adapt health-system innovation in local contexts and limited health human resources due to competing interests across health systems. Both groups emphasized the need for people-centred approaches to establish shared goals and vision, identify gaps and map what has worked to drive health-system innovations, set priorities and discuss how each stakeholder group can contribute to building and reviewing implementation considerations such as resources and funding related for the adaptation and uptake of health-system innovations. Conclusions The findings provide insight for ongoing efforts to improve the development, implementation and evaluation efforts to enhance and harness health-system innovation to strengthen health systems in Canada. Collaboration from within and between governments and sectors will ultimately help to increase the value gained from health-system innovations
Exploring Women’s Sexual and Reproductive Health and Gender-Based Violence Needs in Libya: A Multi-Methods Study
No full textIn this dissertation, I explore women’s sexual and reproductive health (SRH) and gender-based violence (GBV) needs in Libya and examine how structural, political, and normative systems shape access to care. Across five thematically distinct but methodologically connected manuscripts, I trace how systemic breakdown, moral gatekeeping, and legal ambiguity regulate women’s bodies, decisions, and movements, and how women navigate, endure, and resist these layered forms of control.
The research draws on a multi-method qualitative study conducted in Tripoli, Benghazi, and Sabha (2023–2025), including a desk review, a service mapping exercise, in-depth interviews and focus group discussions with women, and key informant interviews. Guided by feminist and justice-oriented frameworks, the dissertation foregrounds women’s lived experiences and the everyday governance of care.
Findings show that access to SRH and GBV services in Libya is structured by interlocking systems of governance. Legal codes, service infrastructures, religious ideologies, cultural norms, and social surveillance do not act independently but reinforce one another, producing conditional and selective access to care. This governance framework generates structural vulnerability, where women’s exposure to harm is normalized, their claims to care are mediated through moral judgment and legal ambiguity, and their autonomy is subordinated to family and national honor. Hegemonic masculinities are institutionalized in health, legal, and security systems that sustain male guardianship and control, while women’s bodies are positioned as repositories of collective morality. Yet women actively resist through secrecy, creative adaptation, linguistic reframing, and online solidarity, tactics that undermine structural constraints while exposing women to risk.
This dissertation provides the first in-depth analysis of SRH and GBV services in Libya and argues that meaningful progress requires systemic change, including legal reform, redistribution of decision- making power, and recognition of women’s adaptive strategies as a foundation for more just and
dignified care
Empowering patients in family medicine research: a case-study of patient partner involvement in educational and knowledge translation activities
No full textAbstract Family doctors are encouraged to involve patients as equal partners in research and teaching, but examples of how to do this well are still quite rare. Our project, called the Canadian Primary Care Information Network - Learner and Patient partnership (CPIN-LEAP), set out to show what genuine partnership can look like when developing learning tools for family medicine residents. What we did. Two patient partners sat on the core research team from day one and helped write the successful grant application. Guided by two national engagement frameworks, we invited nine additional patients to co-design the project’s key products. Together, this team created an online learning module that walks residents through six essential communication skills, and they also drafted eye-catching posters and brochures that invite patients to send feedback after a clinic visit. What worked well. Patient partners said they felt respected, influential, and relevant. Their ideas made the resident module more practical, especially the real-life scenarios and plain-language tips, and ensured that the community posters spoke directly to patients’ concerns. The finished module has already been added to the University of Ottawa’s Department of Family Medicine’s Innovation Portal and clinic participants will be encouraged to display the posters and brochures. Challenges we faced. Recruiting enough French-speaking partners proved difficult in our predominantly English-speaking city, even after we specifically targeted Francophone community groups and social-media pages. Coordinating meeting times was another hurdle, because many partners juggled work, caregiving, or appointments; we solved this by mixing in-person workshops, short evening videoconferences with flexible email feedback. Why it matters. Working side by side with patients led to teaching and knowledge translation materials that reflect real-life worries and priorities. The process shows that deep partnership is doable, but only if projects budget for patient time and stay flexible with timelines. This case study illustrates practical processes and lessons learned from patient partner involvement.Plain English summary Family doctors are being encouraged to include patients as equal partners in research and teaching, but practical examples of how to do this are still rare. Our project called the Canadian Primary Care Information Network – Learner and Patient Partnership (CPIN-LEAP) set out to show what genuine partnership can look like in family medicine. From the very beginning, two patient partners helped shape our ideas and write the successful grant. Later, nine more patients joined to co-create the project’s main materials: an online learning module for residents about patient-centred communication, and bilingual community posters and brochures encouraging patients to share feedback about their care and learn how to get involved in research. The educational module helps residents reflect on what matters most to patients in patient-centred communication during clinical encounters, while the community materials invite patients to make their voices heard. Patient partners said they felt respected and valued, and their suggestions made the educational tools more practical and easier to understand. Working in both English and French, we faced challenges finding enough Francophone partners and coordinating schedules, but flexible meeting formats helped overcome these barriers. This experience shows that meaningful patient-researcher partnerships are possible when time, flexibility, and funding are built into the project from the start
Gas-Liquid Flows in Vertical Annular Channels
No full textExperimental and computational studies of air bubbles rising in stagnant and flowing water in vertical annular channels for the full range of eccentricity (e) were performed. The experiments were conducted in a specially designed experimental facility and two video cameras were used to record the flow from two directions perpendicular to each other. Analysis of the videos and use of computer-aided-design (CAD) software provided measurements of the velocity, length, shape, volume and angle of wrap of Taylor bubbles in stagnant water for 0 ≤ e ≤ 1. The bubble velocity increased with increasing eccentricity from the concentric value (e = 0), peaked for e ≈ 0.2 - 0.3 and then decreased with further increasing eccentricity, reaching a minimum value for e = 1. The maximum angle of wrap decreased monotonically with increasing eccentricity in its full range. The same observations were also made in our numerical simulations. On the other hand, the bubble length and volume were essentially insensitive to eccentricity.
Isolated Taylor bubbles in upward flowing water were found to be slimmer and had a smaller maximum angle of wrap than bubbles with comparable lengths rising in stagnant water. Four regimes were identified visually in upwards air-water flows in annular channels, namely, bubbly, cap-bubbly, slug and churn flows. A flow regime map was constructed for the full range of eccentricity. The transition regions of the map were shifted with changes in eccentricity in the range e = 0.3 - 0.7. Oscillatory cross-flows, attributed to gap vortex streets, were visually identified in the recordings of bubbly, cap-bubbly and slug flow regimes for e = 0.5 and 0.7. The numerical simulations also provided evidence for the presence of gap vortex streets in gas-liquid flows in highly eccentric annular channels. These vortex streets are deemed to be a main source of the eccentricity effects
Embeddable Temporal Road-User Detection from Radar: A Hybrid CNN-MetaFormer Approach
No full textThanks to significant breakthroughs in millimeter-wave radar technology, deep learning architectures, and edge computing capabilities, the pursuit of robust all-weather perception systems for autonomous vehicles has intensified. With various environmental challenges and safety-critical scenarios demanding reliable object detection, researchers are addressing fundamental limitations in sensor-based perception systems. One of the most pressing challenges is achieving accurate road-user detection using automotive radar while maintaining computational efficiency for embedded deployment. Given that modern vehicles require real-time processing to operate on limited computational resources, this thesis presents a hybrid deep learning framework that leverages temporal radar data through a novel CNN-MetaFormer architecture to perform efficient detection and classification of dynamic road users.
We provide a comprehensive analysis of traditional radar processing methods and their evolution toward deep learning approaches, examining both convolutional-based and transformer-based architectures for radar object detection. We also thoroughly investigate temporal modeling strategies and sensor-aware design principles specific to radar data characteristics. Furthermore, we present detailed development of our proposed CompactRADNet architecture that processes sequences of range-azimuth radar frames, introducing the Adaptive Quadratic ReLU (AQR) activation function and radar aware, multipart loss function . Our extensive experiments on the CRUW dataset demonstrate superior performance over state-of-the-art methods. The real-world deployment demonstrates the framework's implementation feasibility, highlighting the impact of hybrid architectural design, temporal sequence optimization, radar-specific adaptations, and the critical balance between detection accuracy and computational efficiency in automotive radar perception systems