140816 research outputs found
Sort by
All For One and One For Gall: Getting Kids Excited About Plant-Insect Interactions
No full textInsect biodiversity is on a decline globally. In order to care about insects, the population must know about them. Creating interest starts with young children and students, using age-appropriate resources and lesson plans. Ensuring that future generations are curious about entomology and not driven away from it is integral to building a future where the public protects the environment. I will add to the available tools by creating my own non-fiction children’s book on under-explored entomology topics
University of Missouri–St. Louis: Program Profile
No full textThe University of Missouri–St. Louis (UMSL) is a large public research institution located in a diverse metropolitan region. The university has approximately 16,000 students, including over 9,000 on-campus undergraduates; the student-to-faculty ratio is 19:1. In fall 2023, on-campus minority enrollment was 40%, and 75% of undergraduates received aid. UMSL is part of the University of Missouri land-grant system and is classified as RU-H (high research activity) and as a Community Engaged Campus by Carnegie rankings
East Carolina University: Program Profile
No full textThe mission of East Carolina University (ECU), which is located in Greenville, North Carolina, is to be a national model for student success, public service, and regional transformation. As of fall 2023, ECU serves over 26,000 students and offers 85 bachelor’s degree programs, 68 master’s programs, 13 doctoral degree programs, and various other certificate and advanced programs. Current enrollment includes nearly 20,000 full-time students (17,104 undergraduates, 2,700 graduate students, 591 in the School of Medicine, and 207 in the School of Dental Medicine). Almost 90% of these students are from North Carolina, with the remaining students coming to ECU from 46 states, Washington, D.C., and 70 countries. Students from underrepresented groups comprise 31% of the population, and over 40% pursue STEM/Health Care degrees. The ECU student-faculty ratio is 18:1, with approximately 2,000 faculty
Rapid State-of-Health Estimation of Batteries Using Machine Learning with Limited Eearly-discharge Voltage Data
No full textThe utilization of lithium-ion batteries has been rapidly expanding across diverse sectors, including electric transportation, stationary energy storage systems, and the built environment. Ensuring a high level of reliability in these applications is essential, as the performance and safety of such systems depend strongly on the accurate assessment of the battery’s State of Health (SOH). Conventional SOH estimation techniques—often based on complex electrochemical models or extensive laboratory testing—tend to require a large number of measurements, advanced instrumentation, and high computational cost. These factors make them impractical for large-scale deployment or real-time monitoring. This study introduces a simplified machine-learning-based approach for fast and reliable estimation of battery SOH using a limited set of easily measurable parameters. The proposed method leverages both simulated data generated from physics-based battery models and experimental data obtained from real lithium-ion cells. The model is trained and validated using standard machine learning algorithms optimized for regression accuracy and generalization across different operating conditions. Results demonstrate that the proposed machine learning model achieves high predictive accuracy with significantly reduced data and computational requirements compared to conventional methods. The approach eliminates the necessity for complex electrochemical simulations or continuous parameter extraction, thereby reducing implementation costs and time. The developed framework has the potential to serve as a low-cost, high-speed alternative to commercial SOH estimation tools, enabling widespread application in battery management systems, grid-level monitoring, and predictive maintenance of energy storage assets.
Advisor: Moe Alahma
LLM-Assisted CWE Identification, Severity Assessment, and Vulnerability Description Generation
No full textIdentifying the underlying weakness types and assessing their severity using CWE and CVSS standards are critical steps in software vulnerability management. While timely assessment of vulnerabilities mitigates the impact of severe security incidents, automating joint CWE identification and severity assessment remains challenging due to the heterogeneity of vulnerabilities across different code granularities and programming languages. In addition, generating vulnerability descriptions is often time-consuming, as it requires extensive manual review, validation, and writing by security experts.
In this thesis, we leverage the capabilities of Large Language Models (LLMs) to automate the identification of CWE identifiers and the assessment of their severity using vulnerability descriptions and corresponding vulnerable code at varying levels of granularity. We further extend our approach to automate the process of generating vulnerability descriptions by incorporating vulnerable code along with the identified vulnerability types. Our evaluations employ quantitative and qualitative metrics, robustness checks against potential noise, and manual analysis of the generated artifacts. Our results indicate that fine-tuning LLMs yields consistent improvements across all granularity levels, achieving over a threefold increase in CWE identification, a 17.68% improvement in severity assessment, and a 21.03% improvement in joint prediction over the baseline, while improving the automatic generation of descriptions that more closely mirror NVD entries. These findings highlight the potential of LLM-based approaches in the joint CWE identification and vulnerability assessment across different code granularities, thereby enhancing automated software vulnerability management.
Advisor: Rahul Purandar
Assessment of Microbial Quality and Safety, Including Aflatoxin B1, in Buckwheat Flour and Groats
No full textBuckwheat (Fagopyrum esculentum), a gluten-free pseudocereal, has gained popularity in global food markets for its nutritional value and functional properties. However, despite its health-promoting potential, buckwheat is vulnerable to microbial contamination and mycotoxin accumulation during cultivation, processing, and storage. These factors present significant food safety concerns that remain understudied, particularly in North America.
This study investigated the microbiological quality and mycotoxin contamination of buckwheat flour and groats collected from different suppliers via online sourcing. Samples were collected in two consecutive years (2024-2025). Microbiological quality, including aerobic plate counts (APC), Enterobacteriaceae, coliforms, E. coli, yeasts, and molds, was evaluated using culture-based methods. Significant contamination was observed in buckwheat flour compared to groats. Samples with elevated Enterobacteriaceae counts were subsequently tested for the presence of Salmonella. This study demonstrates that microbial contamination occurs at various stages of buckwheat production, with significant contamination occurring during processing. Furthermore, it highlights the importance of improved handling and processing practices to ensure the safety and quality of buckwheat products.
This study also investigated the occurrence of aflatoxin B1 (AFB1) in buckwheat flour and groat samples. The study focused on developing and validating methods for quantifying aflatoxin B1 in collected buckwheat samples. Fluorometry and high-performance liquid chromatography (HPLC) were optimized and compared, with HPLC demonstrating superior accuracy, precision, recovery (77–99%), and low matrix effect (0.8%) across spiking levels. The validated HPLC method was used to screen commercial buckwheat samples for AFB1 contamination. The findings demonstrate the significance of processing methods and storage conditions in reducing mycotoxin risks in buckwheat-based products.
Advisor: Andréia Bianchin
Brands & Virtual Personas: How Virtual Influencer Types Impact Brand Awareness
No full textAs social media continues to shape how consumers interact with brands, virtual influencers (VIs) have emerged as a new type of digital celebrity. While previous research has explored how VIs contribute to brand image and awareness, there has been limited attention on how different types of VIs, both anime-like and human-like, affect consumer perceptions across different markets. This study examines how these two types of VIs impact relatability, aspiration, and brand awareness, with a focus on the differences between luxury and mass market. Using a within-subjects experimental design, this study recruited 574 participants with social media experience using MTurk. The survey exposed participants to two types of VIs to explore how consumers’ relatable or aspirational perceptions may impact brand awareness. The results of this study indicate that anime-like and human-like VIs were perceived as relatable and aspirational figures. Additionally, perceptions of relatability were found to play a significant role in increasing brand awareness for both luxury and mass market. In contrast, perceptions of aspiration did not show statistically significant support for brand awareness in either the luxury or mass market.
The results aim to enhance the understanding of VI effectiveness in digital marketing and provide practical insights for brands looking to strategically employ VIs based on their target market.Advisor: Jennifer Johnson Jorgense
Examining Trauma and Substance Use Co-morbidity in African Americans
No full textChildhood maltreatment (CM) is a profound public health crisis that significantly elevates the risk for substance use disorders (SUDs), including alcohol and opioid misuse. Adolescents exposed to CM are especially vulnerable, often turning to substances as maladaptive coping mechanisms during critical developmental periods. African American youth bear this disproportionate burden, yet research addressing their unique experiences remains strikingly limited. Furthermore, while the co-occurrence of trauma-related disorders and substance use is well documented, racial and ethnic minority groups continue to be underrepresented in this literature. Emotion regulation difficulties are increasingly recognized as a mechanism linking PTSD symptoms with substance use, yet their influence on treatment engagement is poorly understood.
This dissertation draws on three complementary studies to illuminate the pathways connecting trauma, substance use, and treatment outcomes among African Americans. The first study, a systematic review encompassing more than 14,000 participants, synthesized evidence on the association between CM and alcohol use in African American adolescents. A clear and consistent positive association emerged, underscoring the urgent need for prevention strategies that are both culturally responsive and developmentally sensitive. The second study, a meta-analysis of nine empirical investigations, examined the relationship between PTSD or trauma exposure and opioid use, with attention to racial comparisons. Results revealed a robust overall association and subtle racial differences, suggesting disparities that warrant deeper investigation. The third study analyzed data from African American men (n = 239) in residential SUD treatment to test whether emotion regulation difficulties mediated the link between PTSD symptoms and treatment duration. Counter to expectations, men with greater difficulties in emotional clarity remained in treatment longer when experiencing elevated PTSD symptoms, suggesting that structured treatment environments may buffer the impact of emotional dysregulation.
Together, these studies offer novel insights into the intersection of trauma, emotion regulation, and substance use among African Americans. They identify emotion regulation as a pivotal factor shaping treatment trajectories, highlight enduring gaps in racially inclusive research, and emphasize the critical importance of culturally competent, trauma-informed care. The findings call for prevention and intervention strategies that move beyond one-size-fits-all models, advancing equity and improving outcomes for African American communities disproportionately affected by trauma and addiction.
Advisor: Dennis McMchargu
Realism, Truth, and Commitment: An Exposition and Critique of Michael Devitt\u27s Constitutive Account of Realism
No full textMichael Devitt argues in Realism and Truth that positive semantic issues are not constitutive of realism concerning the mostly impersonal external world. This is motivated by his second maxim which prescribes that the metaphysical (ontological) issue of realism is to be (sharply) distinguished from any (positive) semantic issue. Devitt’s argument is that his view of realism, entitled ‘Realism’, doesn’t entail and isn’t entailed by any doctrine of truth.
This dissertation presents a critique of the argument indicated above and, thereby, Devitt’s second maxim. I present Devitt’s account of Realism and his overview of theories of truth. I next present Devitt’s argument for why Realism isn’t entailed by and doesn’t entail any positive doctrine of truth. Using Devitt’s discussion of deflationary and correspondence theories of truth, I develop a view which I entitle ‘Mere-Deflationism’. I argue that Mere-Deflationism is a semantic doctrine and that Realism entails it. I conclude that Realism is an adequate form of external world realism only if it is not distinct from any semantic issue.
I expound a semantic objection to Realism from nonfactualism that Devitt considers. Concerning the existential statements which are partly constitutive of Realism, one can assert these statements and yet not be ontologically committed to instances of kinds indicated by the predicate-like terms when these terms are taken descriptively. This enables antirealists to affirm Realism by embracing a nonfactualist-like semantic theory. Devitt responds to this possibility by developing a non-semantic criterion for determining ontological commitment. I argue that this non-semantic criterion doesn’t alleviate the concern and is inadequate. I conclude that an antirealist can adhere to Devitt’s doctrine Realism. From this I argue that Realism is not an adequate form of external world realism. Again, I derive the conclusion that Realism is an adequate form of external world realism only if it isn’t distinct from any semantic issue.
Advisor: Jennifer McKitric