Mason Journals (George Mason Univ.)
Not a member yet
3256 research outputs found
Sort by
Quantitative analysis of feedback received from Claude 3.5 Sonnet on mathematical programming problems using a multi-dimensional rubric framework
No full textLarge Language Model-based auto-graders are emerging as a paramount educational technology tool. However, one must ensure their feedback is accurate, beneficial, and consistent before integrating them into education. Previous studies have shown their auto-grading efficiency in mathematics and science and the importance of prompt engineering, but they focus less on understanding mathematical programming problems. A pre-experiment revealed that Claude 3.5 Sonnet produced the most consistent and accurate feedback for such problems compared to other models, like Microsoft Copilot and Meta Llama 3. In this experiment, four researchers solved five Project Euler problems of increasing difficulty (5-25%) in 40 minutes each, asking for graded feedback from Claude in a revision cycle. Claude graded their solutions using a 22-criteria rubric and a prompt-engineered template designed for consistency. The researchers compared pre/post-revision scores across rubric categories and problem difficulties. The results showed an average improvement of 17.5 points. An ANOVA test resulted in a p-value of 7.36x10-33 for the increase in points. The categories with the most improvement on average included time complexity (+25.45%), efficiency (+22.59%), and edge case handling (+22%), with the least improved being naming conventions (+0.83%). Claude’s feedback was most effective for problems with 10-15% difficulty. Our findings show that LLMs have the potential to greatly increase grades for code by providing feedback. Further research should compare the effectiveness of LLM and human grader feedback with students in an undergraduate-level course, or use models specific to the field with a more diverse dataset
Developing An Artificial Intelligence (AI) Agent for Early Elementary Education Using Dialectical Materialism
No full textArtificial Intelligence (AI) is becoming an essential tool in education, providing innovative solutions for personalized learning. However, early elementary education still needs AI agents to meet young children's unique cognitive and social needs. This study proposes the development of an AI agent based on the principles of dialectical materialism, which highlights the dynamic and interactive nature of learning.
This research aims to create an AI system incorporating socio-cultural and historical contexts into its algorithms, making it more adaptable and responsive to students' educational needs. The methodology involves designing an AI that uses qualitative analysis of classroom interactions and theoretical models to predict and adapt to individual learning trajectories.
This AI agent could significantly enhance engagement and learning effectiveness in early elementary students, especially in literacy and numeracy. By providing contextually relevant feedback and adapting to individual learning paths, the AI is expected to support academic performance and social development.
This research suggests that using dialectical materialism in AI design can lead to more effective educational tools that improve learning outcomes and support holistic child development. This proposal lays the groundwork for future studies to test and refine the AI agent in real-world educational settings
The Anti-Inflammatory Effect of Cannabinoids on HIV-1-Infected Monocytes
No full textThe Human Immunodeficiency Virus, type 1 (HIV-1), is a human retrovirus that targets the host immune cells, inhibiting the immune response and incorporating itself into the genome of immune cells4. The number of estimated infections of HIV-1 worldwide is approximately 37 million people, predominantly in areas such as Sub-Saharan Africa, Southern Asia, and South America4. Currently, the treatment options mainly focus on combination Antiretroviral therapy (cART), which induces the virus into latency, resulting in undetectable levels of viral load. However, during latency, the HIV-1 infected cells continuously undergo low-level transcription4. This low-level transcription prompts residual viral particles to activate the NF-kB pathway, responsible for early regulation of the adaptive immune response, involving the production of cytokines and resulting in chronic inflammation of the central nervous system leading to HIV-associated neurocognitive disorder (HAND)3. During this experiment, the HIV-1-infected and uninfected monocyte cell lines U1 and U937, respectively, were treated with increasing concentrations of HU308 and CBD, both compounds have been shown to reduce inflammation in the CNS, likely due to the inhibition of NLRP3 inflammasome activity and thus reducing inflammatory cytokine release and pyroptosis1,2. Here we looked at the expression of NLRP3 inflammasome components and subsequent cytokine release, in both the whole cell extract and cell supernatants, derived from U1 HIV-1 infected cells after treatment with CBD and HU308.
References:
Blevins, H. M., Xu, Y., Biby, S., & Zhang, S. (2022). The NLRP3 Inflammasome Pathway: A Review of Mechanisms and Inhibitors for the Treatment of Inflammatory Diseases. Frontiers in aging neuroscience, 14, 879021. https://doi.org/10.3389/fnagi.2022.879021
Dopkins, N., Miranda, K., Wilson, K., Holloman, B. L., Nagarkatti, P., & Nagarkatti, M. (2022). Effects of Orally Administered Cannabidiol on Neuroinflammation and Intestinal Inflammation in the Attenuation of Experimental Autoimmune Encephalomyelitis. Journal of neuroimmune pharmacology: the official journal of the Society on NeuroImmune Pharmacology, 17(1-2), 15–32. https://doi.org/10.1007/s11481-021-10023-6
Mu, W., Patankar, V., Kitchen, S., & Zhen, A. (2024). Examining Chronic Inflammation, Immune Metabolism, and T Cell Dysfunction in HIV Infection. Viruses, 16(2), 219. https://doi.org/10.3390/v16020219
Ventura J. D. (2020). Human Immunodeficiency Virus 1 (HIV-1): Viral Latency, the Reservoir, and the Cure. The Yale journal of biology and medicine, 93(4), 549–560
Constructing a Comprehensive Dataset of Reproducible Flaky Tests
No full textRegression testing, an important part of software development, is essential for ensuring that new software changes don't break existing features and ensuring software reliability. This process relies on the assumption that test results are consistent: the same test should always produce the same outcome for the same version of code. However, flaky tests, which non deterministically pass or fail, reduce software reliability. These inconsistent tests hinder the ability to accurately identify software regression issues and slow down development. While datasets of flaky tests exist, it can be difficult to reproduce these tests because of their inherent nondeterminism and existing datasets mostly contains two categories of flaky tests. To address this problem, we aim to create a more comprehensive dataset of reproducible flaky tests including other categories as well. To address this, we study the issue reports and flaky test fixes from various open-source projects to create a script that can reliably reproduce some flaky test failures as described in these reports. Our dataset includes flaky test categories hardly found in prior datasets and helps contribute to a deeper understanding of these test failures. By creating this dataset, we hope to facilitate the development of advanced techniques for flaky test detection and resolution, ultimately improving software reliability
Flakewatch: Automated Categorization and Detection of Flaky Tests At Scale
No full textAutomated testing is essential for developing stable and effective software, as it enables developers to implement changes confidently while minimizing the risk of regressions in functionality. However, automated testing suffers from flaky tests, i.e. tests that both pass and fail nondeterministically on the same version of code. Flaky tests can be categorized based on their source of nondeterminism, including dependency on test order, false assumptions about their execution environment, and other sources of differing behavior between executions. While multiple techniques for detecting specific types of flaky tests exist, no tool can combine these detectors to perform categorization. Additionally, while large datasets of flaky tests exist, these tests are often outdated, and thus it can be difficult to reproduce their failures. We present Flakewatch, an automated tool to create a live dataset of categorized flaky tests, based on the output of existing detection tools. Flakewatch’s combination of multiple detection tools’ output helps to avoid false negative detections, where a true flaky test is incorrectly marked as not flaky. Given a project, Flakewatch will continuously pull for commits, then run detectors on tests found to be added or modified. Flakewatch then updates a live database of active flaky tests, including any that failed in continuous integration. Flakewatch contributes to furthering the development of automated flaky test detection tools and may help improve understanding of root causes and solutions to flaky tests
The Effects of Wildfire Aerosols on Phytoplankton Populations in the North Pacific Ocean
No full textAerosols released by wildfires can have varying effects on phytoplankton populations. Although wildfire smoke can block sunlight and inhibit phytoplankton growth, wildfire aerosols also deposit phosphorus, nitrogen, and iron onto nutrient-limited oceans, providing a major food source for phytoplankton. Phytoplankton are critical to oceanic food webs, and fluctuations in phytoplankton populations could bring drastic changes to oceanic ecosystems. In this study, I explore relationships between aerosols emitted by intense California wildfires in 2020 and phytoplankton in the nearby North Pacific Ocean during this time. I analyzed chlorophyll-a, Photosynthetically Available Radiation (PAR), Aerosol Optical Depth (AOD), Particulate Matter under 2.5 micrometers (PM2.5), and nutrient data from satellites by creating climatologies, maps, and time series of these variables from 2003-2020. For July through November, PM2.5 emissions from wildfires and chlorophyll-a amount (an indicator of phytoplankton) were both highest in 2020, during the Giga Fire period, in the North Pacific compared to 18 years of previous wildfire seasons. Visual results display significant, positive deviations from climatological means in chlorophyll-a of up to 170% in September of 2020. This reveals anomalously high chlorophyll-a levels in the North Pacific Ocean following periods of wildfires, implying a large increase in phytoplankton populations. The existence of positive correlations between wildfire emissions and phytoplankton growth are demonstrated to be possible in this study, leaving immense implications for the dynamics of global aquatic ecosystems
Analyzing the Impact of Different Funding and Governance Models on a OSS Project’s Success
No full textWith the rise of blockchain technology, especially Bitcoin and Ethereum, many projects aim to enhance the blockchain environment by improving security and facilitating cryptocurrency exchanges. Open-Source Software (OSS), allowing public use and modification, is crucial in these advancements. GitHub, a popular OSS platform, provides valuable insights into project activity and community participation. This study investigates factors contributing to OSS project success, particularly the impact of funding sources and organizational structures.Data from 660 GitHub OSS projects, spanning from 2013 to the present, were analyzed using SQL and Python scripts. Key metrics such as commit counts, watches, and pull requests were collected to gauge activity levels. Additionally, a qualitative study categorized each project based on functionality, governance, and funding models, covering projects like cryptocurrency wallets, decentralized applications (Dapps), Layer 2 networks, and tokens.Preliminary results indicate that different funding and governance models significantly affect project activity on GitHub. These findings suggest that the structure and financial backing of OSS projects are vital to their development and success. Although data analysis is ongoing, the study aims to provide insights to help developers and stakeholders optimize project design and funding strategies, fostering sustainable growth and success in the OSS community.
This abstract is part of a collection in which the overarching large project under Dr. Jiasun Li was subdivided into discrete critical tasks that were carried out by multiple individuals or smaller teams. Abstracts in this collection read similarly given the shared project goals, but represent distinct tasks completed by the abstract authors towards finalizing the described analysis
A Multivariate Analysis of Blockchain-based Open Source Software (OSS) on Github
No full textOpen Source Software (OSS) solutions are the backbone of nearly every technological development in the 21st century. Valued at $8.8 trillion, different OSS have been implemented in around 96% of commercial programs produced and distributed to the public. Blockchain-based technologies have also seen a huge spike in demand, compounding at an annual rate of 68%. However, an increasing number of these blockchain projects have gone private. By 2017, only 8% of blockchain-based OSS were actively maintained from the previous year on Github. This problem culminated in this 2-stage analysis of over 600 Github-based blockchain projects. During the first stage, Google’s BigQuery enabled tracking of Github activity from 2013-2024, including watches, commits, and other forms of user contributions to the different communities. In the second stage, these OSS were later categorized into 10+ project types including: crypto exchanges, Layer-1 protocols, Layer-2 networks, decentralized applications, tokens, and Web 2.0 business among others. Further sorting was performed by funding models, governance modes, and governance tokens if applicable. The goal of this analysis is to recognize patterns within the Github community data to determine success factors responsible for active maintenance of a project. This study found a majority of projects to be decentralized applications, yet there was a significant minority of other projects. Other factors were also found in high correlation with active maintenance of an OSS. This analysis hopes to provide insight into elements influencing declining blockchain-based OSS and offer guidance for future OSS initiatives.
This abstract is part of a collection in which the overarching large project under Dr. Jiasun Li was subdivided into discrete critical tasks that were carried out by multiple individuals or smaller teams. Abstracts in this collection read similarly given the shared project goals, but represent distinct tasks completed by the abstract authors towards finalizing the described analysis
Optimizing Governance In Blockchain Open-Source Software: Trends and Insights Across The GitHub Timeline
No full textGitHub is a web platform that enables programmers to share their code to fellow developers, facilitating seamless collaboration. Many GitHub projects are open-source software (OSS), whose source code is readily available to the public. OSS is essential to decentralization in blockchain, fostering innovation towards more resilient and secure blockchains by shifting control from a central entity to a distributed network. These projects utilize a wide variety of funding sources and governance models. This study aims to investigate activity on such projects and find trends with their respective funding and governance models. Our data was found on GitHub Archive, a project that streams and archives public events on GitHub such as issues, logins, or commits. We first ran an SQL script to access all of the GitHub activities for 600+ OSS projects. This script was run on BigQuery, a Google Cloud platform utilizing Google’s powerful processing infrastructure to gather data at high speeds, convenient for large-scale data collection of millions of records. Next, our team conducted thorough research on these projects and the organizations that governed them, recording key details such as their funding, governance, and associated cryptocurrencies. An analysis of a smaller set of data showed that projects governed by non-profits were typically funded by crypto token sales or crowdfunding/donations. This suggests that some trend(s) indeed may exist between funding, governance, and GitHub activity. By comparing these key details with the frequency of various GitHub events, this analysis can provide unique insights into the operational frameworks that yield the most success for various organizations within the blockchain ecosystem.
This abstract is part of a collection in which the overarching large project under Dr. Jiasun Li was subdivided into discrete critical tasks that were carried out by multiple individuals or smaller teams. Abstracts in this collection read similarly given the shared project goals, but represent distinct tasks completed by the abstract authors towards finalizing the described analysis
Secretory Mitophagy: A new target for ovarian cancer drug resistance
No full textOvarian cancer, which forms in the fallopian tubes, is characterized by asymptomatic early stages, resulting in late detection, a poor prognosis, and a high fatality rate. Additionally, ovarian cancer has higher rates of drug resistance. Recent research has highlighted the importance of mitochondrial dynamics, particularly fusion and fission, in ovarian cancer development and drug resistance. Mitochondrial fusion combines two mitochondria to reduce stress and share mitochondrial DNA, whereas fission separates mitochondria and allows for the elimination of damaged components via mitophagy. We discovered a unique mechanism called secretory mitophagy, in which damaged mitochondrial proteins and components are ejected via extracellular vesicles and mitochondrial fission. Our research identified the FIS1 protein as a critical fission initiator. Cancer cells enhance mitophagy in response to oxidative stress. Unexpectedly, ovarian cancer cells have enlarged mitochondria, suggesting enhanced fusion. We propose that increased mitochondrial fusion leads to chemoresistance in ovarian cancer. By inhibiting FIS1, we expect enhanced mitochondrial fusion, connecting fusion to drug resistance. We further hypothesize that secretory mitophagy allows ovarian cancer cells to withstand significant oxidative stress. To test these assumptions, we used siRNA against FIS1, treated cells with mitophagy inducers (Carbonyl cyanide m-chlorophenyl hydrazone (CCCP)) and lysosomal inhibitors (Bafilomycin A), and then measured cell viability and toxicity. Our tests included western blots for PINK1, FIS1, CD81, TOM20, and DRP1. Our data suggests mitochondrial fusion and secretory mitophagy play critical roles in ovarian cancer drug resistance and survival pathways, respectively. Targeting these mechanisms may provide new treatment options for this aggressive disease