Carleton University Institutional Repository
Not a member yet
20816 research outputs found
Sort by
Nature, Nurture, and Healing: Designing Holistic Non-Clinical Healing Spaces for Cancer Patients
No full textThis thesis proposes a non-clinical space designed to enhance the well-being of cancer patients, integrating nature to support emotional, psychological, and physical recovery. Moving beyond traditional Western hospitals, the project introduces biophilic principles to create an environment that combines therapeutic gardens, sensory pathways, and opportunities for patients to grow their food. Inspired by Maggie’s Centres and Toronto’s Casey House, this non-clinical healing space aims to meet the holistic needs of cancer patients, addressing the gaps in current urban healthcare design. By blending communal areas for connection, private rooms for reflection, and nature-inspired zones for relaxation, the design offers a supportive retreat for those facing the challenges of cancer treatment. Case studies and direct observations from healing-focused facilities will guide the project, providing insights to shape a future where architectural design is vital in nurturing health and well-being beyond medical treatment, aiming to reshape urban healing environments
Lingering Atmospheres: Collaborations between Architecture, Craft, Music, Landscape, and Film in Peter Zumthor’s Gesamtkunstwerk
Get PDFThis dissertation explores the connections between multidisciplinary design and multisensory experience by investigating the collaborations between architecture, craft, music, landscape, and film in Swiss Architect Peter Zumthor’s Gesamtkunstwerk (Total Work of Art). If sense perception is understood as fundamentally multisensory in nature, and the experience of architecture is dependent on the people, situations, and events it houses, how might we re-examine occularcentric and unidisciplinary narratives of architectural history? I focus on four temporary exhibitions that Zumthor co-curated: Swiss Sound Box (Hannover Expo 2000), Peter Zumthor: Buildings and Projects 1986-2007 (Kunsthaus Bregenz 2007), Hortus Conclusus (Serpentine Gallery Pavilion 2011), and Peter Zumthor: Dear to Me (Kunsthaus Bregenz 2017). These projects make explicit the centrality of cross-disciplinary collaboration and cross-modal perception in Zumthor’s practice. They are lingering atmospheres: designed as Gesamtkunstwerk and realised through collaborations with musicians, literary critics and writers, chefs and hospitality experts, artists, filmmakers, landscape designers, and engineers. They shift focus away from architecture as a hermetically sealed discipline toward an architecture that is inherently intertwined with other disciplines. By investigating Zumthor’s images as imaginative gestures that engage with other disciplines as well as past and future atmospheres, this research offers new insight on Zumthor’s creative process and works. I differentiate between three kinds of images in relation to Zumthor’s practice: ‘beforeimages’ as concrete dreams in the embodied mind that catalyse each project, ‘images’ as physical drawings, models, and words that embody and translate ‘beforeimages,’ and ‘afterimages’ as photographs, audiovisual recordings, and multimedia representations that extend atmospheres. I argue that Zumthor co-creates atmospheres by intertwining four types of imagination: material, sonic, landscape, and filmic. As the title to the 2017 publication Dear to Me: Peter Zumthor in Conversation suggests, it is not the Peter Zumthor as sole proprietor, but rather Peter Zumthor “in conversation with” that fosters the most evocative atmospheres. I aim to elucidate how his design process foregrounds the interdependent nature of architecture and its surrounding disciplines and reveal the importance of ‘afterimages’ as synesthetic embodiments that enable co-created atmospheres to linger past the moment of experience
Energy Recovery and Treatment of Brewery Wastewater in a Bioelectrochemically Enhanced Anaerobic Digestion System
No full textThis thesis focused on the development of a bioelectrochemically enhanced anaerobic digestion (BEAD) system to enhance energy recovery (in form of methane gas) from brewery wastewater at room temperature. The research was conducted in three stages. First, it assessed the impact of varying volatile fatty acid concentrations in the feed with gradual transition to brewery wastewater. Over 90% COD removal at all feeds demonstrated the system’s ability in handling simple and complex feeds such as brewery wastewater. Second, BEAD’s performance was evaluated on increasing organic loading rates (OLRs) from 1 to 7 g sCOD/L.d, achieving over 92% COD removal and a high methane yield of 0.30 LCH4/g sCODRemoved at an OLR of 5 g sCOD/L.d. Lastly, comparative assessment of the BEAD’s performance under control conditions indicated up to 30% higher methane production in comparison to traditional AD operation
Learning Robust Graph Neural Networks with Limited Supervision
Get PDFGraph Neural Networks (GNNs) have demonstrated significant success in various graph-related tasks. However, their performance is highly dependent on the availability of sufficient, class-balanced annotated data and an accurate graph structure. GNNs are vulnerable to substantial performance degradation when confronted with limited annotated samples, class imbalance, or noisy graph structures. First, a Dual GNN learning framework is introduced for semi-supervised node classification under limited supervision. The framework comprises two GNN-based node prediction modules: a primary module, which uses the input graph structure to induce standard node embeddings and predictions, and an auxiliary module, which leverages spectral clustering to construct a new graph structure and learn new node embeddings and predictions. These modules collaborate to enable end-to-end learning of discriminative node representations. Next, a graph augmentation method called Graph Dual Mixup (GDM) is proposed for graph classification with scarce labels. GDM employs a graph structural auto-encoder to learn structural embeddings, applying mixup in the learned structural embedding space to generate new graph structures. Additionally, mixup is applied to input node features, generating node features for new graph instances. Together, these generated node features and graph structures contribute new graphs that increase the size and diversity of the labeled dataset, improving classification performance. To enhance robustness against adversarial attacks on graph structures, an Efficient Low-Rank GNN is introduced. This approach learns robust low-rank sparse graph structures through a two-stage process. In the first stage, Singular Value Decomposition is used to estimate a low-rank approximation of the graph structure. This estimate is refined in the second stage by jointly learning a low-rank sparse graph structure alongside the GNN model, resulting in enhanced robustness. Finally, to address class-imbalanced node classification, a Unified GNN Learning (Uni-GNN) framework is proposed. Uni-GNN integrates structural and semantic connectivity representations to extend the propagation of embeddings to non-adjacent structural neighbors and semantically similar nodes, facilitating the diffusion of discriminative information. Additionally, it incorporates a balanced pseudo-label generation mechanism to improve the representations of minority classes. Comprehensive evaluations on benchmark datasets demonstrate that these methods outperform state-of-the-art approaches and hold significant potential for advancing GNN learning with limited supervision
Novel Techniques and Applications in Mass Spectrometry Based Lipidomics Analysis
Get PDFMass spectrometry (MS) is one of the most important analytical techniques for research and industry. Advances in MS technology coupled to liquid and gas chromatographic (LC-MS and GC-MS, respectively) techniques have enabled the thorough characterization of complex biological samples. Lipids are an important and chemically diverse group of metabolites that have a wide range of roles in eukaryotic cells. Lipidomics, the large-scale study of lipids in biological systems, is an emerging field of bio-analytical chemistry. MS is the primary technique used to investigate lipidomic dynamics. Lipidomics has existed for more than 2 decades and has made significant contributions to our understanding of health, medicine and biological processes. This thesis primarily focuses on new insights and developments to improve MS-based lipidomics and outlines how they can be used for emerging lipidomics applications. LC-MS fatty acid data extraction (LC-MS FADE) is a novel technique developed to estimate the fatty acid profile in complex biological samples that were previously only accessible using GC-MS instrumentation. A platform-independent in silico reference database of lipid fragmentation spectra chemically derivatizatized lipid classes. Analysis of 3 specific lipid classes revealed that alkylated amine contaminants found in LC-MS grade methanol and isopropanol used as the LC mobile phase caused dramatic effects on their annotation and quantification highlighting the difficulties of reproducibility of lipidomics experiments. All the previously described developments and observations were applied to a novel lipidomics application, the production of lentiviral vectors used for advanced biotherapeutics. In the following dissertation, the first lipidomic analysis of HEK 293T packaging cells producing lentiviral vectors is described, along with the first ever attempt to characterize the lipid composition of clinical lentiviral vectors. Through these findings and detailed methods, future studies aimed at improving vector yield or quality through manipulating lipids may be pursued
Exploring Audience Engagement with Sustainability Influencers: Insights from Two Field Studies
Get PDFInfluencer marketing has become a powerful tool across various industries. Sustainability influencers, who promote eco-friendly lifestyles on social media, have gained significant attention from environmentally conscious audiences. This thesis investigates audience engagement with five categories of sustainability influencers: Zero Waste, Ethical Fashion, DIY and Upcycling, Minimalism, and Vegan influencers. Through two field studies conducted on YouTube, this thesis aims to provide valuable insights into effective audience engagement strategies. Study 1 reveals that higher language arousal, paired with an educational appeal, significantly increases both passive and active engagement. Study 2 uses topic modeling to analyze audience comments to identify key drivers of active engagement, which include influencer expertise, attractiveness, parasocial relationships, and audience motivations such as information seeking, social interaction, and community identification. This thesis contributes practical strategies and academic insights into how sustainability influencers can effectively engage their audiences and encourage sustainable practices