24067 research outputs found
Sort by
Large language model enabled mental health app recommendations using structured datasets
No full textThe increasing use of large language models (LLMs) in mental health support necessitates detailed evaluation of their recommendation capabilities. This study compares four modern LLMs—GPT-4o, Claude 3.5 Sonnet, and dataset-enhanced Gemma 2 and GPT-3.5-Turbo—in recommending mental health applications. We constructed a structured dataset of 55 mental health apps using RoBERTa-based sentiment analysis and keyword similarity scoring, focusing on depression, anxiety, ADHD, and insomnia. Standard LLMs demonstrated inconsistent accuracy and often relied on outdated or generic information. In contrast, our retrieval-augmented generation (RAG) pipeline enabled lower-cost models to achieve up to 55% higher accuracy than baseline models while recommending apps with significantly better user ratings. Dataset-enhanced models maintained perfect accuracy while preserving recommendation diversity and quality. These findings demonstrate that strategically enhanced, cost-effective LLMs can outperform expensive proprietary models in domain-specific applications like mental health resource recommendations, potentially improving accessibility to quality mental health support tools
Next-Gen Donors: Designing a Virtual Reality Blood Donation App for Young Donors
No full textBlood donation is a fundamental procedure in healthcare. Donors can donate blood to help those in need, but sometimes, the process may be timely and uneasy to understand. This results in a shortage of blood donations. In addition to this factor, most donors are over the age of 45 and the percentage of young adult donors ages 17 to 24 have fallen from 13.07% to 7.2% in 2022-2023 according to NHS Blood and Transplant. To encourage young adults to donate blood and establish a simple donation process, we created a blood donation app called,”The VR Blood Donation Company”. The goal of our app is to increase donor engagement, specifically in young adults, improve blood donor management, and facilitate the connection between donors. The app includes features such as virtual reality, scheduling appointments, and communication with clinics. By using frontend programming, HTML and CSS, we made the app visually appealing to the eye and through backend programming, JavaScript, the app is able to store user information and enhance the navigation of the app. Our objective is to make a positive impact on the healthcare industry and promote blood donations. The VR Blood Donation Company makes this objective possible by encouraging young adults through virtual reality and aesthetic designs, simplifying the donation process with straightforward navigation, and ultimately saving those in need
3D Printing Waterproof Geocache Containers
No full textSince the Covid-19 pandemic, universities have struggled with achieving strong student engagement; this has had a number of detrimental effects on the students. One solution which has shown strong results in helping increase student engagement, as well as a wide range of other benefits, is geocaching. In practice, it would work as a system of scavenger hunts around campus. A limiting factor to the number of campuses that can adopt this technology is longevity of the containers designed to be found in the scavenger hunts. Often, they are flooded due to poor waterproofing and their contents ruined, leading many to abandon trying to use them entirely. Our research question focuses on determining how we can design waterproof 3D-printed containers and determining which qualities, such as print settings, shape, and material contribute most to waterproofness. One method used to test waterproofness was found by weighing the containers before and after being submerged in water for a week. The containers were submerged in a bucket and rocks were put in the containers to keep them under water. Another method used was filling the containers with dyed water to see where the water was leaking out. Our results indicate that there are methods we can follow to make PLA printed containers waterproof or at least water-resistant, so that these containers can be used for geocaching purposes. These results are based on different materials, models, settings, and slicing parameters
Exploring Protein Interactions to Understand Gene Regulation by the Drosophila Ecdysone Receptor
No full textThe Ecdysone Receptor (EcR) is a nuclear receptor found in invertebrates, such as drosophila, that regulates gene expression during development and reproduction. While it is known that the ligand binding domain (LBD) of the ecdysone receptor binds some accessory proteins besides the ligand to regulate gene expression, protein-protein interactions with the ligand binding domain of the ecdysone receptor is understudied, and identifying the full counsel of these proteins can give major insights into how the system regulates gene expression. This project uses the Yeast Two-Hybrid (Y2H) assay to investigate novel physical interactions between the LBD of EcR and proteins in a drosophila cDNA library. The LBD is fused to the DNA binding domain of Gal4, while a drosophila library of cDNA is fused to the DNA activation domain of Gal4, creating bait and prey proteins respectively. When bait and prey are transformed into yeast, and if they interact, Gal4 transcription factor is successfully reconstituted which then activates reporter genes resulting in blue colonies and resistance to the antibiotic aureobasidin, enabling the identification of novel protein-protein interactions. A preliminary check for autoactivation was completed by transforming our bait into yeast and performing the assay to ensure that reporter genes are not activated by the LBD in the absence of the DNA activation domain. Y2H controls were also performed; the positive control confirmed interaction between known interacting proteins p53 and T-antigen and the negative control confirmed that known non-interacting proteins do not activate reporter genes. By completing the check for potential autoactivation and positive and negative controls of the Y2H assay, it was determined that the assay is functioning as designed, and further experimentation can be completed. The library screening will be completed to determine proteins that positively interact with the LBD of EcR and potentially aid in gene regulation
GC/MS Analysis of Allergens in Perfumes and Allergen-Protein Interaction via Molecular Docking
Get PDFBeautification products have been around for a very long time, with perfumes being among the oldest forms. Many different ingredients are added to these perfumes, such as essential oils from different plants, to achieve certain desired scents. As the development of new and more beauty products expands, the study of possible unknown or hidden compounds in those products intensifies. We focused on the identification and quantification of allergen compounds, specifically coumarin and benzyl alcohol, present in various perfumes. The interaction between allergen molecules and human leukocyte antigens (HLA) protein after a perfume is sprayed and in contact with human skin is also our research interest. Interested compounds in perfumes and other beauty products can be identified and quantified via chromatographic separation followed by mass spectroscopic analysis (GC/MS). Bromobenzene was selected as the internal standard, and the elution procedure was developed via the Shimadzu QP2010 Plus GCMS system. Both coumarin and benzyl alcohol have been resolved and identified in a 27-minute elution process. The calibration curves with and without internal standards have been constructed with excellent linearity (R2 = 0.999). Both allergens from various perfume samples have been quantified at low ppm levels. The allergen-protein interactions via molecular docking were analyzed via the HDOCK server. Preliminary results show that the unbiased docking (-92.8 kJ/mol) between the HLA-B protein and benzyl alcohol has better binding affinity than the bias docking (-68.33 kJ/mol). Further research will be done with molecular docking for coumarin, and the results will be summarized and presented
Exploring Beneath the Surface: Mapping the Underground with AI and Radar
No full textGround-Penetrating Radar (GPR) is a vital technology for non-invasive subsurface exploration, enabling the identification of buried objects and structural anomalies. A key challenge in interpreting GPR data is the reliable recognition of hyperbolic signatures within noisy B-scans. In this work, we present a deep learning-based framework for automated hyperbola detection that directly learns from real GPR data, eliminating the need for manual feature engineering. Leveraging a carefully curated dataset of annotated radar images, our convolutional neural network model extracts and exploits high-level features to distinguish hyperbolic reflections from clutter and other background interferences. Experimental evaluations demonstrate that our approach achieves superior accuracy and robustness compared to traditional methods, reducing both false positives and missed detections. This advancement not only accelerates GPR data processing but also improves the reliability and scalability of subsurface mapping. By harnessing the power of AI and radar, we provide an automated solution that can be easily integrated into existing workflows, paving the way for more efficient geological surveys, infrastructure assessments, and archaeological explorations
PROTEINS THAT BIND WITH THE LBD OF THE ECDYSONE RECEPTOR RESPONSIBLE FOR GENE REGULATION IN DROSOPHILA
No full textThe Ecdysone Receptor (ECR) is a nuclear receptor in Drosophila regulating gene activity. ECR consists of a ligand binding domain (LBD), which binds the ligand and other proteins, enhancing its ability to regulate gene expression involved in signaling pathways linked to processes such as cell growth, cell differential, and metamorphosis. However, many proteins are still unknown. This research aims to discover new interacting proteins via yeast two-hybrid assay. This assay uses the modular nature of Gal4, a transcription factor, with a DNA-binding domain (BD) and a DNA activation domain (AD). Gal4 AD-prey proteins from a drosophila cDNA library will be screened with Gal4 DNA BD fused to LBD (bait). If bait and any prey proteins interact, the Gal4 transcription factor is reconstituted and reporter genes are activated, as indicated by blue-colored colonies and resistance to antibiotic aureobascidin. So far, yeast transformations, LBD autoactivation, and positive and negative mating controls have been performed. White colonies appeared on all plates; therefore, the yeast transformations were successful. No colonies appeared on the DDO/X/A plate with LBD; therefore, Gal4 DNA BD-LBD does not autoactivate the reporter genes. Blue colonies appeared on the DDO/X/A plate for the positive control, and no colonies appeared for the negative control. All the controls work as expected, so we will proceed with the library screening of potential LBD protein interactors using the yeast two-hybrid assay. Gaining insights into these interactions expands our understanding of the ECR dynamics in Drosophila. This knowledge has potential implications for human health as nuclear receptors are linked to similar cellular activities
Shell Bead Production and the Middle Preclassic Maya
No full textThe ancient Maya are well known as one of the major complex societies in the New World. Within their society, they were known for making many different artisan goods, one of which was the shell bead. It is hypothesized that these beads were not only used for jewelry and personal adornment but also as a form of early currency. The first examples of early shell bead production are found during the Middle Preclassic period (1000-300 BC), these beads are more rudimentary, they lack much sanding, and have rough shapes. As time progressed, they started making these beads all the same size and shape which likely means they had some way to grind all of the shells at the same time to keep them uniform. My research focuses on the shell beads found at Pacbitun, located in west central Belize, where 5,670 beads, 516 chert drills, and large amounts of detritus have been found, indicating that this site likely specialized in creating shell beads. This project uses experimental archaeology to identify the tools and techniques used to create these large quantities of shell beads
AI-Integrated BIM for Net-Zero Construction: Strategies for Efficiency, Waste Reduction, and Worker Safety
No full textThis research aims to explore how AI technology in construction can advance net-zero buildings by improving efficiency, reducing waste, and ensuring ethical decision-making, including worker safety and secure construction management. BIM (Building Information Modeling) software that integrates AI technology can be used as an approach to understanding how construction zones for sustainable buildings maintain a safe workplace for workers and the environment. Because BIM is a foundation of different AI software, using the correct software for construction is also crucial for achieving this goal. This information on how AI technology balances the environment, and decision-making can be gathered through research papers that cover similar findings. These research papers include trends, case studies, and data-driven results that can be used to identify patterns, compare methodologies, and evaluate the effectiveness of AI technology solutions. Since AI technology hasn\u27t been implemented in construction until the early 2000s, we are also gathering information that can highlight the constructive and destructive effects of BIM’s development. This research will mainly present the ethical decision-making of BIM technology and will provide feedback on how we can improve AI-driven project management software tools. The findings of this research are expected to inform future improvements in AI-driven project management tools for BIM technology to incorporate within its system, thereby improving any disruptions in the work area and creating a more sustainable future for the construction industry
Evaluation of Compassion Fatigue and Perceived Organizational Support in Georgia Animal Rescues
No full textAnimal rescue volunteers often face emotionally demanding situations, making them vulnerable to compassion fatigue. Compassion fatigue (CF) combines elements of burnout with secondary traumatic stress and can impact an individual’s physical and mental health. We examined the relationship between animal rescue volunteers’ levels of CF and the degree to which they felt valued and supported by the organizations for which they volunteered (Perceived Organizational Support (POS)). We distributed surveys to 104 animal rescue organizations in Georgia, yielding 259 valid responses. Our sample was majority female (88.4%) and White (91.5%), primarily volunteering with dogs (65.3%) and cats (34.0%). The survey combined a Professional Quality of Life Scale (assessing compassion fatigue through subscales of compassion satisfaction, burnout, and secondary traumatic stress) and a shortened and modified POS scale. Our data violated normality assumptions, so we used Kendall’s tau correlation coefficient to analyze the data. Preliminary analysis revealed that POS was partially related to CF, correlating negatively with burnout (τb= -.230, p \u3c .001) and positively with compassion satisfaction (τb= .323, p \u3c .001). However, the analysis also found that POS did not correlate significantly with secondary traumatic stress, (τb= -.041, p = .353). This study is the first to compare levels of CF and POS in animal rescue volunteers. Organizations can use this information to better support their volunteers, leading to increased volunteer retention over time