33760 research outputs found
Sort by
NEO-FUTURE: Envisioning the Transformation of the Human Environment Through Space Architecture
No full textThis thesis envisions human environments beyond the current Earth-bound limitations of architecture, extending into Low Earth Orbit and Lunar Orbit. It seeks to transform our traditional understanding of architecture and shape a more sustainable future for human habitation. The proposal is a space-based habitat that transcends pure functionality. It fosters a symbiotic relationship between humans and their extraterrestrial environment, addressing not only physiological challenges but also the psychological well-being of its occupants. By deepening our understanding of the human-environment interaction, this research works to create a framework for sustainable habitation beyond Earth while simultaneously enriching our understanding of human centered design and sustainability applicable to terrestrial contexts. The design explores innovative solutions such as artificial gravity mechanisms, sustainable life support systems, community-oriented design, modular assembly, and automated construction. Furthermore, this thesis investigates strategies for creating a nurturing environment in the harsh expanse of space through principles of biophilic design and communal spaces conducive to resilience and social cohesion. By synthesizing insights from various disciplines including architecture, aerospace engineering, psychology, and human physiology, the aim of this research is to act as a first step to propel humanity towards a future where the line between terrestrial and extraterrestrial life begins to dissolve. Ultimately it seeks to catalyze a shift in mindset—one that harmonizes the relationship between our human species, the environments we design, and the cosmos around us
SOMETHING ABOUT ME
No full textSomething About Me explores the resurfacing of fragmented memories as the speaker of the poems reconciles with her past. The collection unfolds in a liminal space between past and present versions of the speaker as she converses with a former self as though through an existential mentoring. In a series of intimate recounts, she confronts her grief as a fatherless daughter, examines moments where the body became a vessel for punishment, and challenges her tumultuous relationships. As a collection, the poems work as a trajectory of catharsis; the speaker purges threats and revelations from adolescence, attempting to free herself from the rejection, shame, and fear that consumed her upbringing
“WALKING DEAD" EN LA ERA DIGITAL: NECROPOLÍTICA Y TESTIMONIOS DE MIGRACIÓN EN TIKTOK (2021-2024)
No full textThis dissertation examines the audiovisual testimonies produced by forcibly displacedmigrants—primarily from Cuba, Venezuela, and Nicaragua—who document and share their
journeys to the United States on TikTok. Through the lens of necropolitics and digital
humanities, it explores how a selection of videos on TikTok—mediated by smartphones,
algorithms, and platform politics—serves as an assertion of existence and collective memory.
While the routes traversed and captured on the videos—through the Darién Gap, across Mexico,
and into the waters of the Río Bravo—are shaped by state-sanctioned violence and border
regimes, they are also marked by solidarity, friendship, and an enduring will to survive. Rather
than focusing solely on death or despair, this research foregrounds the agency of migrants as
self-archivists. Their recordings, raw and immediate, carve space for visibility in a digital
landscape that often erases or distorts their presence. TikTok, while enabling this documentation,
also subjects these narratives to the fleeting rhythms of virality and algorithmic bias. In response,
this dissertation is also accompanied by a living digital archive (website) titled Voices Across
Borders, which preserves fragments of testimonies and resists their disappearance into the scroll.
This work is a gesture of witnessing. It asks what it means to document life at the edge of death,
and how storytelling, no matter how pixelated or precarious, can affirm dignity. It is, moreover, a
meditation on the power of hope that keeps migrants walking
The Communalization of Trauma through the Visual Arts: Images of War in Greece and Etruria in the 6th and 5th centuries BCE
No full textThis thesis explores the cathartic potential of the visual arts through an investigation of images of the war dead on painted vessels and gems from 6th and 5th century BCE Greece and Etruria. My research builds on the large body of work on the cathartic properties of the arts, an idea first proposed in Aristotle’s Poetics and investigated in modern scholarship by authors such as Jonathan Shay. My goal is to frame the ancient visual arts as a mode for cathartic release through the communalization of trauma and grief as a result of war. I investigate imagery on painted vessels, largely from tomb assemblages in Etruria and Athens as well as a group of mostly unprovenanced intaglios and gems attributed to the Etruscans and/or Greeks. This research will demonstrate that, when placed correctly within their social, mythological, and historical contexts, these objects and their imagery of the war dead would have aided in grieving, healing from trauma, and loss of agency in times of war
Exploring the Contextual Influence of a Summer Jobs Program on Youth Arrest: Evidence from New York City
No full textGovernment-sponsored job programs have gained significant attention in recent years, particularly the Summer Youth Employment Programs (SYEPs), which connect young people to government-subsidized jobs during the summer months. Originally designed to improve youth employment, SYEPs have also been proposed as a policy to combat youth crime in the summer when youth typically face increased risks of crime, injury, and victimization (Sepúlveda & Hutton, 2019). Evidence supports this idea, indicating that participation in SYEPs is linked to lower rates of arrest, arraignment, conviction, and incarceration, with effects varying by youth’s race, age, gender, and prior justice system involvement (Gelber et al., 2016; Heller, 2014; Kessler et al., 2022; Modestino, 2019). Prior research suggests that neighborhood context may also condition the effects of SYEP participation on criminal justice outcomes, but this has not been systematically evaluated. This dissertation addresses this gap by examining whether neighborhood factors, such as disadvantage and police stop, question, and frisk (SQF) activity, shape SYEP’s effect on youth arrests. Analyzing data from 129,098 youth aged 16 to 21 to who applied to the New York City SYEP between 2006 to 2008, I find that effects of SYEP on short-term and long-term arrests were nuanced and may differ based on the applicants’ residential context and racial/ethnic identity. The dissertation contributes to the broader discourse on summer job programs and discusses critical considerations for policymakers
TOWARDS EFFECTIVE AND EFFICIENT VIDEO UNDERSTANDING
No full text“If a picture is worth a thousand words, what is a video worth?” Video information, due to its inherent richness and efficiency compared to language, plays a pivotal role in conveying complex information. However, video understanding faces numerous challenges, including selecting informative frames, addressing domain shifts, semantic grounding, reasoning and attention deficits, and significant computational burdens. Recent advancements in computer vision underscore the need to address these challenges through effective and efficient approaches, which are crucial for applications ranging from autonomous systems to human-computer interactions that require high accuracy and low latency. In this dissertation, we address five critical issues to overcome these challenges: dataset development, preprocessing, visual reasoning, multimodal alignment, and computational acceleration.
High-quality datasets serve as the foundational building blocks, providing diverse, comprehensive, and representative data to train models capable of handling real-world complexity. In this dissertation, we proposed METEOR dataset for tailored for autonomous driving applications in dense, heterogeneous, and unstructured traffic scenarios with rare and challenging conditions. Additionally, we developed DAVE, a comprehensive benchmark dataset specifically designed to enhance video understanding research for the safety of vulnerable road users in complex and unpredictable environments. Our analysis revealed substantial shortcomings of current object detection and behavior prediction models when tested against our METEOR and DAVE.
Complementing datasets, for preprocessing, we proposed AZTR incorporates an automatic zooming algorithm
for dynamic target scaling and a temporal reasoning mechanism to accurately capture action
sequences. Furthermore, we introduced MITFAS, an alignment and sampling method based on mutual information specifically designed to address challenges inherent to UAV video action recognition, including varying human resolutions,
large positional changes between frames, and occluded action features.
For visual reasoning,
we introduced SCP, which guides the model to explicitly learn input-invariant (prompt experts) and input-specific (data-dependent) prompt knowledge, effectively capturing discriminative patterns and significantly improving accuracy on challenging datasets. We also developed ICAR, a compatibility learning framework with a novel category-aware Flexible Bidirectional Transformer (FBT), which can effectively generate features across different domains based on visual similarity and complementarity for reasoning tasks.
For multimodal alignment, we proposed ViLA to address both efficient frame sampling and effective cross-modal alignment in a unified way. Finally, we propose Bi-VLM to explore ultra-low precision post-training quantization method to bridge the gap between computational demands and practical limitations. Our method employs a saliency-aware hybrid quantization algorithm combined with a non-uniform model weight partition strategy, significantly reducing computational costs without compromising much overall model performance
STETHOSCOPES TO #SELFCARE: THE TRANSFORMATION OF EXPERTISE DYNAMICS IN DIGITAL ENVIRONMENTS
No full textThis dissertation examines how expertise is communicated, performed, and assessed across different digital environments through three complementary case studies: physicians on Twitter/X, mental health content creators on TikTok, and graduate students assessing AI-generated content. Each study reveals how traditional markers of authority operate when moved into digital spaces with distinct features and affordances.
The research demonstrates that digital platforms do not simply host expertise claims, but actively reshape how expertise can be signaled, validated, and perceived. When traditional markers of professional authority such as credentials, specialized vocabulary, and visual signifiers move online, they undergo transformations specific to each platform's architecture. On Twitter/X, physicians leverage profile features to display credentials while developing content that serves multiple audience segments. TikTok's visual emphasis requires mental health creators to perform expertise through new hybrid forms that combine professional knowledge with platform-native styles. The AI study reveals how graduate students develop mental models for evaluating seemingly authoritative non-human content produced without domain knowledge.
The central contribution of this work is documenting how digital environments transform traditional mechanisms of expertise validation, creating conditions where expertise communication adapts to platform-specific environments. These findings extend beyond individual platforms to inform our understanding of how expertise signals operate across different digital contexts and how users approach the assessment of competing knowledge claims in increasingly AI-mediated spaces.
The dissertation provides insight into how different platform affordances, professional domains, and audience expectations create varied conditions for expertise communication while revealing common patterns in how authority adapts to digital constraints
Co-Designing the Co-Design Process: Formulating Participatory Studies and Co-Creating Robotic Interventions with Autistic Children and Their Families
No full textWhen serving as participants, autistic children and their families are all too often brought in at the end of the study to evaluate robotic interventions with the premise of support. Even when co-design approaches are employed, its involvement takes place after the problem formulation. Within this thesis we take a different approach: co-designing the co-design process. This study involves the key stakeholders of autistic children, their families, and clinical psychologists, in not only the creation of the robotic intervention, but also in the foundational design of the study itself. This two-tiered engagement underscores a comprehensive participatory approach, ensuring that both the research framework and the resulting solutions are deeply informed by stakeholder experience and insight. Through iterative brainstorming sessions, study material design, and pilot co-design sessions, we formulate a robotic intervention in the form of a game, created to practice frustration management, and support cognitive flexibility. The contributions of this thesis are mainly empirical, we provide study materials that have been co-designed with our stakeholders. However, there are methodological underpinnings in this work as well, which highlight the novelty of this two-tiered approach to co-design in this space
Navigating Stress and Coping Strategies in Chinese American Families Amidst the COVID-19 Pandemic
No full textThis qualitative study examines the stressful experience of Chinese American families during the COVID-19 syndemic by focusing on perceived stress, coping strategies, socialization of coping, and posttraumatic growth. The sample was drawn from a larger longitudinal study of 529 Chinese American parents, which included both surveys and qualitative interviews. In this dissertation, I analyzed data from 47 semi-structured interviews with Chinese American parents of 12- to 18-year-olds conducted between March and May 2021, followed by nine participants in three follow-up focus groups in September 2024. Guided by the grounded theory, the data was analyzed with MAXQDA 24. The findings showed that Chinese American parents experienced well-being concerns, daily disruptions, as well as racial discrimination. Chinese American parents appraised stress by considering environmental, personal, cultural, and identity factors. Social connections emerged as the most commonly reported coping strategy across various stressors. Many Chinese American parents also socialized their children to cope with racial stress by supporting their children to promote social change and embrace their American identity. Despite the ongoing challenges, Chinese American parents reported posttraumatic growth in areas such as renewed perspectives on ways of living, evolved parenting approaches, strengthening interpersonal connections, and making social impacts. A conceptual model was developed based on the results of this study and it highlighted the concept of flourishment, along with the dynamics of navigating multiple identities in the stress and coping process. This model makes a significant contribution to the field, as many similar stress process models fail to fully address the intersectionality of cultures and identities. Implications for mental health practitioners include the need for culturally informed interventions that consider the unique stressors and experience faced by Chinese American families
When News Headlines Go Wrong: An In-depth Analysis and AI-driven Intervention of Misleading News Headlines
No full textMisleading news headlines that distort, exaggerate, or omit information without presenting outright falsehoods pose a persistent challenge in the digital news ecosystem. These headlines often exploit commercial and algorithmic pressures, taking advantage of limited reader attention and heuristic processing. Despite their widespread impact, misleading headlines have received limited in-depth investigation in both misinformation research and HCI. This dissertation investigates the issue through a multi-method, three-part inquiry: examining human perceptions and correction practices (Project 1), testing the behavioral effects of headline correction strategies (Project 2), and evaluating large language models' (LLMs) capacity to support editorial reasoning (Project 3).
Project 1 explores how two key stakeholder groups, journalists and news readers, perceive and respond to misleading headlines. Through semi-structured interviews with 12 journalists and 12 readers, the study identifies competing notions of responsibility, with journalists emphasizing audience literacy and readers expecting inherent trustworthiness. The analysis surfaces three key correction strategies that stakeholders independently employ: adding uncertainty cues, restoring critical context, and removing emotional framing. These findings reveal editorial tensions and motivate the need to assess how such strategies function when deployed at scale.
Project 2 builds on these qualitative insights through a between-subjects experiment with 399 participants, testing the effects of the three correction strategies on reader outcomes. The study evaluates six headline versions across engagement, credibility, and interpretation accuracy. Results show that corrections, particularly the removal of emotional language, can significantly enhance perceived credibility and interpretive accuracy without diminishing engagement. The findings challenge the presumed trade-off between truthfulness and reader interest and offer empirical grounding for ethical headline design in journalism and platform interventions.
Project 3 investigates how LLMs such as GPT and Gemini explain misleadingness in headlines under varying levels of annotator agreement. Using a stratified dataset of 60 headlines and explanations generated by three LLMs, the study engages six professional journalists to evaluate explanation quality along editorial dimensions, including correctness, ambiguity awareness, and risk sensitivity. While LLMs align well with human reasoning in high-consensus cases, they often falter in ambiguous ones, failing to surface interpretive complexity or journalistic reasoning. The analysis informs design directions for editorially aligned, expert-in-the-loop AI systems.
Together, the three chapters advance a situated understanding of misleading headlines as a socio-technical problem and offer design-relevant implications for computational journalism, explainable AI, and platform governance. This dissertation highlights the need for editorial transparency, role-aware collaboration, and systems that support nuanced, context-sensitive decision-making