University of Illinois Urbana-Champaign
Illinois Digital Environment for Access to Learning and Scholarship RepositoryNot a member yet
123813 research outputs found
Sort by
Computational and machine learning tools for insights into conjugated materials
No full textSubmission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2025-10-19 without embargo termsThe student, David Friday, accepted the attached license on 2025-04-14 at 11:42.The student, David Friday, submitted this Dissertation for approval on 2025-04-14 at 11:48.This Dissertation was approved for publication on 2025-04-16 at 16:04.DSpace SAF Submission Ingestion Package generated from Vireo submission #21765 on 2025-10-19 at 18:18:14Conjugated materials are a versatile class of electro-active organic materials with applications in energy, health, and computing. Central to improving their performance and enabling the implementation in devices is understanding their electronic properties (e.g. electronic mobility and photo-activity). These properties depend on quantum mechanical (QM) properties traditionally computed via DFT. However, DFT is unable to access the large length scales required to predict morphological-dependent properties (e.g. mobility), and the connection between QM-calculable properties and experimentally relevant molecular properties is not always clear. This work seeks to address these limitations by both developing new computational methods enabling prediction of morphologies and QM-informed electronic properties at experimentally relevant length scales, and methods to discover mechanistic insights from experimental campaigns via machine learning. The application of these methods elucidates novel mechanisms driving electronic mobility and photostability of conjugated materials, enabling their further optimization for future applications
Children of the corn: An exploration of Midwest Latinx identity through the discourse of food
Get PDFSubmission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2025-10-19 without embargo termsThe student, Ana Lisa Eberline, accepted the attached license on 2025-04-16 at 11:33.The student, Ana Lisa Eberline, submitted this Dissertation for approval on 2025-04-16 at 12:09.This Dissertation was approved for publication on 2025-04-16 at 13:38.DSpace SAF Submission Ingestion Package generated from Vireo submission #21795 on 2025-10-19 at 18:18:18The purpose of this research is to explore Latinx experiences working and living in the rural and suburban Midwest. Working to understand how this specific population constructs and/or maintains their cultural identity across predominantly White landscapes. By narrowing the scope and focusing on Latinxs within the food industry, I am underscoring a setting that historically has served as a gateway for Latinx labor in the U.S. and remains a site where culture is shared within the community. Given the predominance of Mexican and Latinx restaurants scattered across Midwest highways and backroads, food is a space where “ethnic” identities are to some degree accepted and maintained, even in the smallest towns. As a Latina who grew up in the rural Midwest, I have often wondered how my personal experiences fit into the larger identity of Latinidad. Through the exploration of participant narratives and everyday discourse, this study aims to critically examine how the imagery of food, discussions of food, and memory of eating reflects the complexity Latinidad in connection to place. Observing the ways Midwest Latinx identity specifically navigates the multiplicity of their identity across time and context. Additionally, this research acknowledges existing stereotypes of Latinx people, and explores how intersections of race, gender, socioeconomic status, citizenship, religion, and language play into structures of power and privilege
From geometry to graphs and back: Geometric range searching and algorithms in structured graphs
Get PDFSubmission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2025-10-19 without embargo termsThe student, Da Wei Zheng, accepted the attached license on 2025-04-18 at 17:13.The student, Da Wei Zheng, submitted this Dissertation for approval on 2025-04-18 at 17:31.This Dissertation was approved for publication on 2025-04-22 at 08:59.DSpace SAF Submission Ingestion Package generated from Vireo submission #21840 on 2025-10-19 at 18:18:27The thesis has two parts. The first part is on geometric data structures related to range searching with linear ranges whose boundaries consists of lines or line segments, and semi-algebraic ranges whose boundaries can be described by algebraic equations. These data structures are fundamental components of geometric algorithms such as for nearest neighbors, Euclidean minimum spanning tree, and computing Voronoi diagrams. These problems reduce to solving an offline data structure problem involving n points and n ranges. In this regime, we show that an algorithm running in O(n^4/3) time exists, which match existing lower bounds. We also study the extremes of the data structure tradeoff for when the number of points is much larger than the number of ranges, and vice versa, and provide improved bounds for simplex ranges in these extremes. Additionally, we consider the dual problem of range stabbing, where a set of ranges are preprocessed, and we wish to report information about the ranges stabbed by a query point. For semi-algebraic ranges, we give improved data structure tradeoffs for range stabbing that avoid dependence in the exponent on the complexity of the semi-algebraic ranges. In addition, the aforementioned results all lead to improved algorithms for handling line segment (or algebraic arc) intersections. The second part of the thesis is focused on developing algorithms for structured classes of graphs such as planar graphs, minor-free graphs, and geometric intersection graphs. There are many algorithmic questions that are provably difficult to do in general graphs, but are possible if we assume some additional structural properties. For planar graphs, we show that in the massively parallel model of computation, problems like computing connected components, approximate shortest paths, and approximate flows and cuts, can be computed in constantly many rounds. For minor-free graph families, we use techniques involving VC- dimension to show that subquadratic time algorithms for computing diameter, eccentricity, and Weiner index exist, as well as subquadratic space distance oracles with logarithmic query time exist. These results hold for unweighted directed graphs, and our distance oracle results even extend to directed graphs with real edge weights. For geometric intersection graphs, we show that a truly subquadratic time algorithm for computing the diameter in unit-disk graph exists. This was an open question that was repeatedly asked, in light of conditional lower bounds that rule out subquadratic time algorithms for intersection graphs of spheres in three dimensions, or even for unit line segments in the plane. The algorithm we present involves an involved combination of techniques including: low diameter decompositions of graphs, VC-dimension, and geometric range searching data structures
Algorithmic aspects of connectivity and density in graphs and hypergraphs
Get PDFSubmission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2025-10-19 without embargo termsThe student, Shubhang Kulkarni, accepted the attached license on 2025-04-20 at 11:53.The student, Shubhang Kulkarni, submitted this Dissertation for approval on 2025-04-21 at 16:17.This Dissertation was approved for publication on 2025-04-24 at 09:38.DSpace SAF Submission Ingestion Package generated from Vireo submission #21842 on 2025-10-19 at 18:18:27This thesis investigates algorithmic problems in combinatorial optimization centered around modifying a given network—via deletion, augmentation, or reconfiguration—to achieve tar- get connectivity or density bounds. We study associated optimization problems on graphs, hypergraphs, and submodular functions. Our main contributions are: • Hypergraph Splitting-off. We introduce a splitting-off operation in hypergraphs and prove an analogue of Mader’s theorem: in every hypergraph, a vertex can be re- moved via splitting-off while preserving all pairwise edge-connectivities. We give a strongly polynomial-time algorithm in weighted hypergraphs, with applications including a constructive characterization of k-hyperedge-connected hypergraphs and an alternate proof of an approximate min-max relation for Steiner rooted-connected orientations. Our framework extends to symmetric skew-supermodular functions. • Hypergraph Connectivity Augmentation. We study augmentation to achieve target pairwise connectivities subject to vertex degree constraints. We give a strongly polynomial time algorithm, improving prior pseudo-polynomial results. Our method extends to generating near-uniform hypergraphs, simultaneously augmenting two hypergraphs, and to covering skew-supermodular functions. Applications include strongly polyno- mial time algorithms for node-to-area and mixed-hypergraph connectivity augmentation. • Graph Density Deletion. We study vertex deletion on graphs where the goal is to delete a minimum-cost subset of vertices so that the densest subgraph has density at most a given target ρ. When ρ ≤ 1, this problem is 2-approximable. In contrast, we show logarithmic hardness of approximation for all fixed integers ρ > 1. We also study a generalization to monotone supermodular functions, show approximation equivalence to Submodular Set Cover, and design bicriteria approximation algorithms. • Feedback Vertex Set (FVS) and Pseudoforest Deletion Set (PFDS). We undertake a polyhedral study of FVS and PFDS, special cases of graph density deletion for appropriate ρ ≤ 1. Both problems are 2-approximable, but lacked polynomial-time solvable LP relaxations with matching approximation guarantees. We establish the first such LP formulations for both problems. For PFDS, we resolve a question of Bodlaender, Ono and Otachi by exhibiting an extreme point property of an associated polytope
Questions around symplectic capacities
Get PDFSubmission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2025-10-19 without embargo termsThe student, Jonghyeon Ahn, accepted the attached license on 2025-04-22 at 14:38.The student, Jonghyeon Ahn, submitted this Dissertation for approval on 2025-04-22 at 14:42.This Dissertation was approved for publication on 2025-04-25 at 10:27.DSpace SAF Submission Ingestion Package generated from Vireo submission #21895 on 2025-10-19 at 18:18:39One major question in the study of symplectic capacities is the strong Viterbo conjecture: all normalized symplectic capacities agree on convex domains. This conjecture has spurred extensive research and numerous interesting results about this conjecture have been established. In this thesis, we contribute to this area by focusing on the comparison of symplectic capacities. We explore the interplay between symplectic and convex geometry. More specifically, we study convex bodies in \RR^{2n} with the property that their mean width, one of the most fundamental measurements in convex geometry, cannot be decreased by the action of natural classes of symplectomorphisms. Our main finding is that toric symmetry is a preferred feature of convex bodies that are in optimal symplectic position with respect to the mean width. We then study an -equivariant version of Varolgunes' relative symplectic cohomology. As an application, we construct a relative version of the Gutt-Hutchings capacities and a relative version of the symplectic (co)homology capacity. We show that these relative symplectic capacities can detect the diplaceability and the heaviness of compact subsets of a symplectic manifold. We also compare the first relative Gutt-Hutchings capacity and the relative symplectic (co)homology capacity and prove that they are equal to each other under a natural convexity assumption
A pragmatic and human-centered approach to promoting software accessibility: Design, education, governance
Get PDFSubmission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2025-10-19 without embargo termsThe student, Kyrie Zhou, accepted the attached license on 2025-04-22 at 16:43.The student, Kyrie Zhou, submitted this Dissertation for approval on 2025-04-22 at 16:56.This Dissertation was approved for publication on 2025-04-23 at 19:05.DSpace SAF Submission Ingestion Package generated from Vireo submission #21899 on 2025-10-19 at 18:18:39In this dissertation, I explore three approaches to software accessibility: design, education, and governance. I aim to facilitate a discussion toward a pragmatic and integrated approach to accessibility. In the design chapter, I approached cancer survivors and spotlighted their lived experiences and challenges with software. A survey study, in-depth semi structured interviews, and a diary study were conducted to capture the multi faceted software experiences of cancer survivors. Cancer survivors may have multiple disabilities, including visual impairments, dexterity impairments/neuropathy, and chemo brain, which is unique to this population. These impairments led to difficulty in using software, which is a fundamental way for cancer survivors to search for health-related information and socialize with others. Based on the results, I proposed simple yet practical accessibility guidelines to accommodate the software needs of cancer survivors. For example, bulleted content, highlighting key information, and simplified User Interface (UI) designs make navigation easier for people with chemo brain. Such guidelines can be easily incorporated into software design and development practices. In the education chapter, I validated the possibility of teaching software accessibility before post-secondary education. Accessibility education was found to be limited in the high school curriculum, which is a missed opportunity to equip the next-generation software designers with the mindset of accessible design. I used a set of disability simulation games to successfully enhance the knowledge, awareness, and empathy of high school students regarding accessibility. Games and gamified approaches helped engage students in learning advanced computing concepts. In the governance chapter, I examined the status quo of governance approaches to improving software accessibility. I analyzed 39 software accessibility laws in 32 countries or regions to understand how they did well and where they fell short. One notable gap is that software accessibility laws have very few rules, degrading their enforceability. The laws are often written vaguely and are not friendly to law novices. Better laws stand out with concrete technical recommendations and fine-grained implementations, i.e., distinguishing between websites and mobile applications. In synthesizing the commonalities and differences between the three approaches to software accessibility, I argue that each approach has advantages and limitations — they ultimately complement each other. The design approach may best exercise a direct impact on software practitioners’ mindsets and practices if conveyed in an approachable manner. Yet designers may find limited incentives to follow guidelines outlined and promoted by accessibility researchers, which are not always in line with corporate priorities such as shipping products fast. The education approach can best prepare next-generation software practitioners. However, the long-term and actual effects of educational interventions are unpredictable and need further interrogation. The governance approach, in theory, has the power to force companies to comply with accessibility requirements. Yet, it barely works out in reality because of the ambiguous wording in accessibility laws and a lack of enforceability arising from scarce rules. I argue that an integrated approach combining design, education, and governance interventions should be used to address software accessibility issues and create a more inclusive software ecosystem
Resilience on campus: exploring the retention narratives of Black male professionals in higher education
Get PDFSubmission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2025-10-19 without embargo termsThe student, Christopher Pevey Harry, accepted the attached license on 2025-04-24 at 13:43.The student, Christopher Pevey Harry, submitted this Dissertation for approval on 2025-04-24 at 14:04.This Dissertation was approved for publication on 2025-04-24 at 15:13.DSpace SAF Submission Ingestion Package generated from Vireo submission #21958 on 2025-10-19 at 18:19:01Higher Education is complex, problematic, and embedded with racism and inequities for students and staff. This dissertation examines resilience on campus through the retention narratives of Black male professionals in Higher Education (HE). It scrutinizes the role that HE plays in perpetuating microaggressions and racism and how it upholds the status quo while claiming to be a great equalizer for those who choose to further their education. It chronicles the counterstories of 7 Black men who work in academic affairs leadership roles at predominantly White institutions to understand the innate or learned resilience techniques that they deploy while traversing microaggressions and racism. Using individual qualitative interviews, which were transcribed and coded, with the analysis triangulated with that of an external research auditor to ensure reliability and validity, this study uncovered that many Black men tackle imposter syndrome as they are often the only Black person on their leadership team. Additionally, it analyzes the notion of unseen and uncompensated labor, the extra work these men take on to overprove that they are worthy of the roles they are hired for, which their White counterparts are not expected to do. It highlights the benefits of mentorship and shows the empowering aspects of embracing Blackness while leading as resilience tools to combat microaggressions and racist attacks on their leadership. It positions both negative and positive lived experiences as leadership development tools and identifies leading collaboratively with compassion as an effective leadership style, a key resilience strategy for Black men. Furthermore, this study uncovers the complexity of the intersectional identities of Black male leaders, prompting some to deploy gender fluidity in their leadership to combat the societal stereotypes and political attacks associated with their Black, male, or queer identities. The findings position Black male leaders as key resources for paving the way for other Black leaders and students, especially those leaders who utilized the grit they had learned during their own first-generation student experiences. In doing so, this study calls HE, allies, and those who work with Black male leaders to understand their roles in listening to and learning from the varied retention stories of those who choose to remain in predominantly White institutions in an effort to better support and advocate for Black men in leadership
Beyond cravings: A phenomenological exploration of relief among individuals with substance use disorders on GLP-1 medications
Get PDFSubmission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2025-10-19 without embargo termsThe student, Jennifer Evans, accepted the attached license on 2025-04-29 at 08:18.The student, Jennifer Evans, submitted this Dissertation for approval on 2025-04-29 at 08:24.This Dissertation was approved for publication on 2025-04-30 at 20:19.DSpace SAF Submission Ingestion Package generated from Vireo submission #22066 on 2025-10-19 at 18:19:34This qualitative, phenomenological study investigated the lived experiences of individuals prescribed GLP-1 medications (e.g., semaglutide, tirzepatide) for weight loss while concurrently living with a substance use disorder. With the growing popularity of GLP-1 medications, which are prescribed to one in eight adults in the United States, this study addressed a critical gap in research by exploring the dual health challenges faced by individuals managing obesity and substance use disorders simultaneously. Using in-depth interviews with ten adults living in the United States (n=10) and interpretive phenomenological analysis, the study examined how participants navigated these intersecting conditions and the impact of treatment on their well-being. The essence of the phenomenon was encapsulated by six key themes: profound relief from cravings, as participants described a significant reduction in substance use compulsions after starting GLP-1 medications; a history of repeated attempts at behavior change, with many recounting cycles of returns to use/binge eating and recommitment prior to finding stability with medication; a connection between substance use and eating habits, as participants reflected on deeply ingrained patterns of reward-seeking and self-soothing through both substances and food; a disconnect from traditional recovery spaces, with some feeling that their experience with GLP-1 medications positioned them outside conventional addiction treatment narratives; a commitment to long-term change, as participants expressed a newfound sense of agency in managing their health; and reconciling perceptions of past and present “self”, as participants grappled with shifting identities in response to their evolving health and recovery journeys. Additionally, medication shortages emerged as a significant barrier, leading to distress and concerns about sustained recovery and weight management. These findings provide novel insights into the impact of GLP-1 medications on individuals with co-occurring obesity and substance use disorders. The study contributes to the theoretical foundations of social work related to integrated health care and underscores the need for more inclusive, compassionate, and accessible treatment approaches. By highlighting the lived experiences of this population, this research informs policy, clinical practice, and future studies aimed at addressing the intersection of substance use, obesity, and chronic disease management
Engineering extracellular vesicles and immune cell homing materials for next generation cancer immunotherapy
No full textSubmission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2027-05-01The student, Rimsha Bhatta, accepted the attached license on 2025-02-17 at 14:42.The student, Rimsha Bhatta, submitted this Dissertation for approval on 2025-02-17 at 14:47.This Dissertation was approved for publication on 2025-02-21 at 10:38.DSpace SAF Submission Ingestion Package generated from Vireo submission #21643 on 2025-10-19 at 19:14:15My Ph.D. research focuses on developing innovative technologies to enable precise and effective modulation of dendritic cells (DCs), to advance cancer immunotherapies that elicit durable cytotoxic T lymphocyte (CTL) responses and robust antitumor efficacy. My work spans two key material platforms: nanosized extracellular vesicles (EVs) and macroscale immune cell-homing biomaterials. For EVs, I developed a novel metabolic tagging technology that installs chemical tags (e.g., azido groups) onto the surface of cell-secreted exosomes, enabling the conjugation of immunomodulatory agents through efficient click chemistry. This approach allows for the functionalization and modulation of EVs to enhance their processing by DCs, leading to improved T-cell priming and antitumor activity. For immune cell-homing biomaterials, I designed a prototype cancer immunotherapy platform using DC-homing macroporous hydrogels with tunable pore size and mechanical properties. Together, these technologies aim to provide tailored solutions for enhancing immune responses and overcoming challenges in cancer immunotherapy. Chapter I introduces the foundational concepts of this research and outlines the current challenges and limitations in the field of cancer immunotherapy. Chapter II explores metabolic tagging and targeting strategies for extracellular vesicles (EVs), focusing on the development of next-generation EV vaccines. By conjugating adjuvants (e.g., CpG) to EVs, we demonstrate the induction of robust cytotoxic T lymphocyte (CTL) responses and enhanced antitumor efficacy. Building on this, Chapter III investigates the potential of antibody-conjugated tumor EVs (e.g., anti-DEC205) as therapeutic cancer vaccines, improving the targeting efficiency and activation of dendritic cells (DCs). Chapter IV extends this work by leveraging chemically tagged exosomes to develop exosome-based hydrogels, which function as potent in situ depot vaccines for sustained immune modulation. In parallel, Chapter V delves into the rational design of immune cell-homing macroporous materials, which preferentially recruit specific immune cells (e.g., DCs) to the material site in vivo, serving as a prototype for cancer immunotherapy. Finally, Chapter VI highlights the ability of these immune cell-homing materials to recruit and modulate T cells in vivo, supporting the development of advanced T cell therapies. Together, these chapters present a comprehensive approach to advancing cancer immunotherapy through innovative biomaterial and EV-based strategies
Towards transit data accessibility: Large language models and software tools for GTFS
No full textSubmission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2027-05-01The student, Saipraneeth Devunuri, accepted the attached license on 2025-03-03 at 16:40.The student, Saipraneeth Devunuri, submitted this Dissertation for approval on 2025-03-03 at 16:50.This Dissertation was approved for publication on 2025-03-06 at 10:23.DSpace SAF Submission Ingestion Package generated from Vireo submission #21658 on 2025-10-19 at 19:14:20In an era characterized by data-driven decision-making, the General Transit Feed Specification (GTFS) has emerged as a global standard for publishing public transit data, enabling unprecedented transparency and accessibility. Despite its widespread adoption, extracting and analyzing transit data from GTFS remains challenging due to its complexity, optional components, and varying agency adherence to the standard. This dissertation addresses these challenges by proposing new tools and methods that make transit data more accessible using software and large language model (LLM)-based techniques. The dissertation begins with a systematic survey of errors in GTFS data across 632 US transit feeds. Approximately 21% of the feeds contain at least one error. The analysis identifies the most common issues, with errors related to the optional "shape_dist_traveled" field accounting for the majority, and fare-related discrepancies forming a secondary cluster. The analysis also demonstrates the limits of identifying errors programmatically, showing that manual inspection is necessary to catch some of the most severe errors. Subsequently, this dissertation addresses the absence of tools for calculating bus stop spacings from GTFS feeds by introducing "gtfs-segments," a Python package that computes summary statistics and visualizes spacing distributions. In addition, it establishes terminology and various weighting schemes for calculating stop spacing statistics. Using "gtfs-segments," stop spacings were computed for 539 U.S. transit providers and 83 Canadian providers, while detailed statistics were produced for 30 U.S. providers, 10 Canadian providers, and a sample of 38 international providers. The analysis shows that different weighting schemes yield distinct "average" spacing values on both a hypothetical sample network and actual transit networks. Notably, the weighted spacings in the U.S. and Canada are narrower than those observed in other regions, yet remain broader than what references in the literature suggest from anecdotal evidence. GTFS data is intricate, comprising over 20 interlinked files with 250+ attributes, each having a description, presence condition, and data type. This dissertation investigates the potential of LLMs in extracting information from GTFS feeds by introducing the "GTFS Semantics" and "GTFS Retrieval" benchmarks to evaluate their comprehension and retrieval capabilities. Benchmarking ChatGPT (GPT-3.5 Turbo and GPT-4) reveals that LLMs exhibit a reasonable understanding of GTFS semantics and can perform "simple" extraction tasks by generating Python code. However, they are prone to hallucinations, particularly in distinguishing attribute-file associations and enumerated attribute types. Furthermore, this leads to poor performance on "complex" tasks that involve multiple files and attributes. The culmination of this dissertation is the creation of "TransitGPT," a chatbot that leverages LLMs to answer natural language queries about GTFS data, such as "What is the longest bus route in Chicago?" TransitGPT helps guide the LLM to generate Python code that extracts and manipulates relevant transit data, which is then executed on a server hosting the GTFS feeds. This framework supports a wide range of tasks—including data retrieval, calculations, and interactive visualizations—without requiring users to have extensive knowledge of GTFS or programming. The LLMs are guided entirely by prompts (through prompt engineering techniques) without the need for fine-tuning or direct access to the feeds, allowing any LLM to serve as a drop-in replacement. Evaluations using GPT-4o and Claude-3.5-Sonnet on a benchmark dataset of 100 tasks demonstrate that TransitGPT significantly enhances the accessibility and usability of transit data, empowering planners, researchers, and the public with an intuitive interface for complex data analysis