Digital Commons Network^™

Autonomous Deficiency Detection And Vision-Language Summarization For Underground Infrastructure On Embedded Edge Systems, Johny Lopez

LSU New Orleans Theses and Dissertations

Aging underground infrastructure poses significant risks to public health and environmental safety, yet structural condition assessment remains bottlenecked by labor-intensive manual CCTV inspections. This thesis proposes a comprehensive algorithmic framework enabling fully autonomous, real-time deficiency detection, geometric assessment, and natural language reporting on resource- constrained edge computing platforms. Three core components address this challenge. First, RAPID-SCAN, a novel semantic segmentation architecture utilizing a Dynamic Feature Pyramid Network and Channel-Spatial Attention, achieves real-time, pixel-precise defect localization with dramatically reduced parameters. Second, an Edge-Optimized Vision-Language Model pipeline employing LoRA and 4-bit QLoRA quantization compresses Phi-3.5 for local deployment, en- abling autonomous technical …

Improving Online Political Discussion With Automated Bot Intervention, Bethanie E. Hackett

Departmental Honors & Graduate Capstone Projects

The quality of political discussions occurring on online platforms or social media sites has been deemed quite poor. To address this issue, I investigated whether a Large Language Model (LLM) can be used to promote civil and productive political discussions by identifying and responding to unproductive dialogue. I fine-tuned an existing LLM to detect elements of problematic dialogue, namely misinformation, misrepresentation of sources, logical fallacies, bias, and toxic language, and then respond in a corrective yet non-confrontational manner. The resulting model is referred to as FroBot and was evaluated through an experiment in which a human participant was placed in …

Barriers To Climate Change And Sustainability Action: Nursing Education And Practice, Dawn Marie Smith

Dissertations

Climate change is one of the most pressing public health emergencies of our time and nurses can have a great impact in their current practice and in the education of future nurses (The Alliance of Nurses for Healthy Environments, n.d.; American Nurses Association, 2023; Health Care without Harm, 2025). Deaths due to rising temperatures, vector-borne illness, and food insecurity related to drought and extreme weather are on the rise (WHO, 2024). It has been estimated that globally over 250,000 additional deaths will be attributed to climate related effects between 2030 and 2050 (Watts et al., 2020; WHO, 2023).

A primary …

Student Programming Behavior With And Without Phone Notification Suppression, Gavin T. Eddington

All Graduate Theses and Dissertations, Fall 2023 to Present

Many students work on programming assignments while receiving notifications from their phones, such as text messages or social media alerts. These notifications can interrupt focus and make it harder to stay engaged with a task. This study examines whether silencing phone notifications helps students stay more focused while programming. 

We collected data from students in an introductory computer science course while they worked on programming assignments. Students completed some assignments with notifications silenced and others without. We measured their activity using software that records typing behavior and identifies when students take long pauses, which can indicate interruptions or loss of …

Large Language Models For Introductory Computer Science Education: Content Generation, Intelligent Tutoring, And Learner Modeling, Muhammad Fawad Akbar Khan

All Graduate Theses and Dissertations, Fall 2023 to Present

This dissertation studies how artificial intelligence, especially large language models such as GPT, can help students learn introductory computer programming when the models are used inside a carefully designed learning system. Instead of focusing on AI as a standalone tool, the dissertation follows a connected story: generating learning resources, building a tutoring platform, running a user study, and then analyzing how students behave while they program.

The work first uses prompt engineering to create a large collection of 11,700 Python exercises aligned with introductory computer science topics. Students and instructors then evaluate these exercises to check whether they are clear, …

St-Fmformer: An Autoregressive Generation Framework For Scientific Ensemble Data Predictions, Md Robiul Islam

All Graduate Theses and Dissertations, Fall 2023 to Present

Understanding how physical systems change over time is important in areas such as weather prediction, fluid dynamics, and environmental science. However, accurately predicting future behavior is difficult because these systems are complex and constantly evolving. 

This research develops a deep learning approach to predict how such systems evolve over time. The model learns patterns from past observations and uses them to generate future states step by step. This provides a faster alternative to traditional simulation methods while maintaining strong predictive performance. 

The proposed method focuses on improving the consistency of predictions over time and is designed to work across different …

An Alternative Representation For Temporal Json, Bishal Sarkar

All Graduate Theses and Dissertations, Fall 2023 to Present

JavaScript Object Notation (JSON) is a common format for representing and exchanging data on the web. Most systems only keep the current version of a JSON document, even though, in many situations, it is also important to know how that document changed over time. For example, an application might need to answer questions such as “What did this record look like last week?” or “How has this list grown over the past year?” A simple way to keep this history is to store a full copy of the document every time it changes, but this quickly becomes wasteful, most of …

Gradient Based Optimization Methods For Robust Learning And Biomedical Signal Modeling, Jarrod Mau

All Graduate Theses and Dissertations, Fall 2023 to Present

This dissertation explores how modern artificial intelligence techniques can be used to better understand complex biological data. Specifically, it develops new machine learning based methods and applies them to two important biomedical problems: analyzing brain signals and studying protein behavior.

The first part of the work introduces a new machine learning approach designed to improve how computers classify structured data. Traditional neural networks are powerful but can sometimes generalize poorly. This research proposes a method that combines the flexibility of neural networks with the reliability of ensemble techniques, leading to more robust and accurate predictions across different types of datasets. …

Scalable Roof Polygon Extraction And Geometric Characterization From Remotely-Sensed Data For Snow Load Assessment, Jashon Newlun

All Graduate Theses and Dissertations, Fall 2023 to Present

Heavy snow accumulation on rooftops is a serious structural risk in cold climates, and understanding how much snow builds up on different types of roofs is essential for safe building design. Currently, most data on roof snow loads comes from small, labor-intensive field surveys that cover only a handful of buildings at a time. This results in far too few measurements of buildings to draw confident conclusions about how snow behaves across communities. This thesis develops and demonstrates a new automated approach for measuring roof snow accumulation and extracting key building characteristics across thousands of buildings at once using airborne …

Measurements And Scaling Of Ion Propulsion Impulse During Driven Magnetic Reconnection, Fatima Ebrahimi, Nicholas A. O'Gorman, Kush Maheshwari, Jongsoo Yoo, Alexandre Sainterme, Hantao Ji

Faculty Publications

Impulse scaling during magnetic reconnection, the magnetic energy conversion to kinetic energy, via direct Mach probe measurements in Magnetic Reconnection Experiment is examined. Ion exhaust velocity and impulse scalings with reconnecting magnetic field during the push phase of driven reconnection are presented. The outflows and impulse measurements are compared with global MHD simulations. Both measurements and simulations reveal a favorable scaling, greater than linear, of impulse with reconnecting field. These scaling results establish that magnetic reconnection could be utilized for plasma propulsion.

Augmented Reality In Fashion Retail: A Walmart Unlimited Study, Chloe A. Mcpherson

Apparel Merchandising and Product Development Undergraduate Honors Theses

As technology continues to evolve, augmented reality (AR) has become increasingly common within the retail and fashion industries. This study explored Gen Z consumers’ perceptions of immersive AR shopping experiences through Walmart Unlimited, an interactive digital shopping platform. The purpose of this research was to better understand how younger consumers respond to AR-enhanced shopping environments and whether these technologies influence attitudes toward convenience, engagement, and sustainability in retail.

A quantitative research design was used for this study. Participants completed the Walmart Unlimited shopping experience and then responded to a Qualtrics survey measuring areas such as immersion, satisfaction, ease of use, …

Beyond Morphology: Testing Molecular Assessment Of Macroinvertebrate Communities In Alpine Streams Of The Teton Range, Usa, Joshua Vollmer

All Graduate Theses and Dissertations, Fall 2023 to Present

Alpine streams and the macroinvertebrate communities that reside within them are threatened by the rapidly changing climate. Long-term monitoring of these communities is required to understand how they are changing and to predict their future trajectories. Historically, stream monitoring and assessment has been performed by collecting and identifying individuals using their morphological traits. However, in alpine aquatic ecosystems, distinct macroinvertebrate species often lack the morphological differences that allow them to be distinguished. DNA metabarcoding, an alternative method for identification which uses genetics instead of morphological differences, is not impacted by this issue, and therefore, may be better suited for alpine …

Adversarial Heterogeneous Agent Learning For Robotic Systems: A Framework For Coordinated Competitive Behaviors, Christopher T. Allred

All Graduate Theses and Dissertations, Fall 2023 to Present

Autonomous robot teams must move reliably, explain their actions, and work together—even in changing or adversarial settings. We present a structured, three-stage pathway that builds coordinated team behavior from strong single-robot skills. First, we develop robust legged-robot locomotion and interpretability using only internal actuator signals. From these proprioceptive cues, robots learn to classify terrain and predict short-term power use, enabling energy-aware movement without external sensors. We further analyze learned behaviors with motif discovery to reveal recurring sensor–action patterns, which clarify how agility emerges and guide reward design. Second, we compose these skills into heterogeneous teamwork using centralized training with decentralized …

Sequential Products Of Anaerobic Denitrification In Calcareous And Noncalcareous Soils By Use Of The Gas Chromatograph, Grant S. Cooper

All Graduate Theses and Dissertations, Fall 2023 to Present

The sequential products of anaerobic denitrification were determined on seven Western soils (four alkaline, two acid, and one neutral) by soil and gas analysis. The soils with 1% alfalfa and added KNO3 were incubated at moistures slightly greater than field capacity and with an atmosphere of He. The soils and gases were periodically analyzed and balance sheets prepared. The sequence of NO3-->NO2-->N2O-->N2 operated in all soils. The rates of nitrogen interchanges and maximal amounts of nitrate, N2O2 and N2 were determined. From this data it was postulated that the rate-limiting process for denitrification in acid soils is …

Leveraging The Iron Oxide Archive Of Mineralization, Deformation, And Exhumation Across Geologic Time, Jordan L. Jensen

All Graduate Theses and Dissertations, Fall 2023 to Present

The iron oxide mineral hematite commonly forms from groundwater near Earth’s surface and can preserve a history of fluid flow, fault slip, and landscape evolution. This record can be decoded using (U-Th)/He thermochronology, a dating method based on the temperature-sensitive ingrowth of helium from uranium and thorium decay. Because later heating and natural variation in hematite crystals can produce ambiguous (U-Th)/He results, I integrate dating with tools that reveal differences in hematite chemistry and textures. I apply these techniques to two distinct settings: ancient granitic rocks in Colorado and an active fault zone in the southern San Andreas fault (SSAF) …

Drought-Tolerant Plants Of Great Salt Lake: Determining Dormancy Break, Germination, And Seeding Timing Requirements To Improve Restoration Under Hydrologically Extreme Conditions, Montana Horchler

All Graduate Theses and Dissertations, Fall 2023 to Present

Wetlands are declining worldwide, and those in arid regions are especially vulnerable to climate change, drought, water diversions, and other human pressures. As a result, land managers increasingly rely on active revegetation to re-establish native plant communities and their associated ecosystem functions. Restoration that utilizes seeds as the propagule source is often the most cost-effective way to revegetate large areas, but in highly variable environments such as exist at Great Salt Lake, Utah, drought and flooding regularly limit plant establishment. Restoration practitioners require science-informed strategies to reach their goal of establishing diverse native plant communities with high native species cover. …

Continuous Symmetry Discovery And Enforcement Using Vector Fields With Application To Time Series And Image Data, Benjamin D. Shaw

All Graduate Theses and Dissertations, Fall 2023 to Present

Machine learning powers many technologies we use every day, from image classification in computer vision to analyzing time series data such as sensor readings or financial trends. One way to make these systems smarter and more reliable is by using symmetry, meaning patterns that remain unchanged under certain transformations, such as rotations for images or uniform changes in signal amplitude for time series. Discovering these patterns can improve understanding of data and help build models that perform well in new situations. 

Current methods for finding symmetry often rely on large neural networks and are limited to simple transformations like stretching …

Social And Institutional Factors Influencing Restoration Decisions In Sagebrush Plant Communities In The Great Basin, Carmen Calzado-Martínez

All Graduate Theses and Dissertations, Fall 2023 to Present

Sagebrush landscapes of the Great Basin in the western United States have changed dramatically over the past century. Invasive grasses, repeated wildfires, and past land-use practices have made it increasingly difficult for native plant communities to recover once they are damaged. Most restoration efforts occur after disturbances such as wildfire, when ecosystems may already be severely degraded. An alternative approach is proactive restoration, which aims to strengthen ecosystems before major damage occurs. 

This dissertation examines whether proactive restoration strategies could realistically be used in sagebrush rangelands and what factors influence their adoption. The research combines three complementary approaches. First, interviews …

Counterfactual Explanations For Time Series Through Local Pattern Mining And Generative Models, Omar Bahri

All Graduate Theses and Dissertations, Fall 2023 to Present

Machine learning systems increasingly make decisions that affect people’s lives, from medical diagnoses to loan approvals. When these systems analyze time series data—sequences of measurements collected over time, such as heart rhythms or sensor readings—users need to understand why a particular prediction was made. Counterfactual explanations address this need by answering “what-if” questions: what would need to change in the input for the system to make a different prediction? 

This dissertation develops methods for generating counterfactual explanations specifically designed for time series data. Time series often contain distinctive local patterns—short subsequences that distinguish one class from another. For example, a …

Post-Wildfire Erosion And Biogeochemistry: Integrating Aerial Imagery And Soil Testing To Assess Landscape Recovery, Justin A. Allred

All Graduate Theses and Dissertations, Fall 2023 to Present

After a wildfire, land managers are required to monitor large tracts of land with limited time and budgets. Traditionally, tracking landscape recovery is a time intensive process. This research explored the effectiveness of using drones (UAVs), strategic soil sampling, and computer modeling as additional tools for land managers in order to monitor the recovery more efficiently.

By using drones to collect aerial imagery and using machine learning, plant regrowth was monitored in back-to-back years. The machine learning model performed well at telling broad groups apart (trees vs grass) but it struggled to identify differences between plant species. Because many plants …

Comparing Strategies For Wetland Revegetation In An Uncertain Water Future, Hailey M. Machnikowski

All Graduate Theses and Dissertations, Fall 2023 to Present

Wetlands worldwide are expected to experience more frequent and intense droughts and floods due to climate change. These shifting water conditions introduce challenges to restoring wetland plants, especially using direct seeding. Seeding is cost and labor-efficient compared to planting, but seeds and seedlings are highly vulnerable to water stress, so restoration efforts often fail when conditions are too wet or dry. Understanding how to improve seed-based revegetation under these conditions is increasingly important. Therefore, we sought to understand the response, measured by percent canopy cover or dry biomass weights, of native plants to different durations of drought in greenhouse experiments, …

Advancing Context-Aware Detection Of Socially Harmful Discourse Using Transformer-Based Models, Santosh Chapagain

All Graduate Theses and Dissertations, Fall 2023 to Present

Social media platforms are a central part of modern communication, shaping how people share ideas, build communities, and discuss social issues. While these spaces can support connection and self expression, they also enable the spread of harmful language such as hate speech. At the same time, social media is an important place where members of marginalized communities, including sexual and gender minorities, express stress, discrimination, and emotional challenges in ways that are often indirect and context dependent.

This research examines whether modern artificial intelligence systems can better identify harmful language and expressions of minority stress in online posts. The study …

Influence Of Coterie On Utah Prairie Dog Translocation Success, Bonnie Stokes

All Graduate Theses and Dissertations, Fall 2023 to Present

The Utah prairie dog (Cynomys parvidens) is a threatened species found in southwestern Utah. Utah prairie dogs are small, burrowing squirrels that live in family groups called coteries. Translocation, the intentional capture, movement, and release of animals from one location to another, has been widely used to aid in their recovery; however, post-translocation survival remains low. The primary objective of this study was to evaluate the effect of translocating Utah prairie dogs as intact coteries on post-translocation survival.

This study was conducted within the West Desert Recovery Unit in Iron and Beaver Counties, Utah, during July–October 2023 and …

Rainbow Dominating Sets Of Graphs, Samuel L. Powell

All Graduate Theses and Dissertations, Fall 2023 to Present

The content of this thesis may be compared to a stack of plates with a common design that are broken, one at a time. If the plates are broken into large enough pieces you would be able to choose one fragment from each broken plate to discover the entire design that the plates share. For example, if you were still missing the center of the design after taking a piece from some plates, you could look for the piece of the next broken plate with the region in question.

We study the question of how many broken plates might guarantee …

Visualizing Probabilistic Model Checking: An Interactive Framework For Exploring Ctmc Models, Ishara Mawelle Kankanamge

All Graduate Theses and Dissertations, Fall 2023 to Present

Probabilistic model checking is a critical method for analyzing systems characterized by uncertainty, such as communication protocols, randomized algorithms, and biochemical networks. While formal verification tools provide precise numerical data about these systems, interpreting these results is often limited by large state space, high-dimensional state space and time dependent evolution. Current tools typically output raw numerical data, offering limited support for intuitively understanding the time-dependent behavior of a model. This research presents an interactive visualization framework designed to bridge the gap between complex numerical analysis and human intuition. The framework integrates coordinated visual interfaces, including lower-dimensional state-space projections and synchronized …

How Novices Write Code: Discovering Best Practices, Matt Rau

All Graduate Theses and Dissertations, Fall 2023 to Present

Learning to program is a difficult endeavor, leading to chronically high failure rates in introductory programming courses. One thing that makes teaching programming difficult is that we don’t fully understand what problem solving habits and writing strategies separate successful programmers from struggling ones. Knowing how to teach these habits to a new programmer is an equally difficult challenge,

Studying the way people write code has proved difficult. Until recently, there was very little relevant publicly available data, and no good ways to analyze student programming behavior at a large scale. In this thesis, I address both issues. I publish a …

Post-Quantum Cryptography Encryption Implementation For Messaging App, Callum S. Ward

Theses/Capstones/Creative Projects

This paper and complementary capstone project aim to explore the state of post-quantum cryptography today by defining the algorithms with which quantum computers can decipher modern asymmetric cryptographic algorithms in exponentially accelerated time, exploring national standards body NIST’s recommendations to circumvent these weaknesses with post-quantum solutions, and implementing recommended algorithms in my group’s project for the UNO Computer Science Capstone course, LockTalk. After having decided on ML-KEM for quantum-resistant asymmetric key transfer and AES-256 for symmetric message encryption and decryption, I was able to cryptographically encode messages to obscure their plaintext values from communication interceptions without any discernible increase in …

Blockchain Message Integrity For Messaging App: Python-Based Implementation Of Blockchain-Backed Verification And Tamper Detection, Brendan J. Farrell

Theses/Capstones/Creative Projects

With the rapid development and use of digital communication, the need for maintaining the integrity and authenticity of transmitted information becomes more pressing than ever.  However, existing methods of data exchange have several flaws and drawbacks such as reliance on centralized networks which are subject to manipulation, modifications, and potential failures.  Thus, the current project offers an innovative approach to improving the integrity and detection capabilities of messages in real-time communication platforms using the power of blockchain technology.  The proposed solution uses the inherent features of blockchain-based systems to ensure secure and safe message transmission.

Message Malware: File Vulnerability Scanning In Messaging Apps, Miah Mason

Theses/Capstones/Creative Projects

As messaging platforms become primary communication hubs within highly secure environments, they introduce significant vulnerabilities regarding file sharing with potential malicious content. This research, an honors extension of the Northrop Grumman sponsored Capstone project ‘LockTalk’, evaluates the technical abilities of automated file scanning methods within messaging platforms like Slack, Microsoft Teams, or our own developed platform LockTalk. This paper investigates a variety of scanning methods and their applications, with particular emphasis on comparing the qualities of static scanning, used in signature-based detection, and dynamic scanning with heuristics. Furthermore, this paper investigates the differences and advantages of both End-to-End Encryption (E2EE), …

Using Ai For Data Loss Prevention, Camden A. Wright

Theses/Capstones/Creative Projects

Data Loss Prevention (DLP) systems play a critical role in protecting modern systems that handle sensitive information from both accidental and malicious exposure. Traditional DLP approaches often rely on static rules and methods that can struggle to adapt to complex and evolving data patterns. This paper presents a hybrid DPL system that integrates machine learning-based message classification, rule based policy enforcement, and context-aware access control to improve both detection accuracy and decision reliability. In addition, the system introduces a second stage access control model that evaluates user context, including role of clearance level and job title to determine whether access …