Digital Commons Network^™

Suicidal Ideation Detection On Reddit Using Llm-Annotated Data And Graph Neural Networks, Ikbal Singh Gurdev Singh Dhanjal

Master's Projects

Suicide is the fourth leading cause of death among people aged 15-29. More than 720, 000 people commit suicide every year. During the COVID-19 pandemic, we saw an increase in people seeking out mental health support on anonymous forums like Reddit. These anonymous forums allow people to express their suicidal ideation without judgment and give them a support structure that not everyone has. The aim of this project is to detect suicidal ideation using Reddit. In this work, we propose SIRGEL (Suicidal Ideation on Reddit using Graph Embeddings and LLMs), a dual-pipeline approach that combines large language model (LLM)- based …

Transforming Chatbot Text: A Sequence-To-Sequence Approach To Human-Like Text Generation, Natesh Reddy

Master's Projects

With the advancements in Large Language Models (LLMs) such as ChatGPT, the boundary between text written by a human and that created by AI has become blurred. This poses a threat to systems designed to detect AI-generated content. In this work, we adopt a novel strategy to adversarially transform GPT-generated text using sequence-to-sequence (Seq2Seq) models, to make the text more human-like. We concentrate on improving GPT-generated sentences by including significant linguistic, structural, and semantic components that are typical of human-authored text. The goal is to then use this transformed data to train a more robust detection model. Our experiments show …

Framework For Identity Privacy Through Gender Based Skeletonization, Harrison Hwang

Master's Projects

The protection of one’s privacy and sensitive information is becoming increasingly difficult in the modern age full of surveillance and data collection. Through the use of image based object detection machine learning models trained for human and facial recognition, people can be identified and tracked to a terrifyingly accurate degree. On the other hand, the information present in surveillance media can play a key role in security and law enforcement. This presents a problem of how to preserve key information without compromising the privacy of any individuals present in the video. In this research project, Computer Vision techniques and a …

Rift - Reddit Information Falsity Tagger, Parth Joshi

Master's Projects

Social media platforms such as Reddit are widely used for sharing and consuming information. User-generated content poses a great risk for misinformation creation and dissemination on these platforms. “Fake news”, as it is commonly referred to, has far-reaching social implications, swaying public perception, making political viewpoints more radical, and adversely impacting health decisions. The covariable features that come with fake news make it even harder to detect because it is presented in the form of text, images, videos, and even social interactions. This paper describes a novel method for detecting fake news on Reddit: RIFT, short for Reddit Information Falsity …

Multilingual Sentiment Analysis Using Ensemble Learning, Farhan Ansari

Master's Projects

The widespread use of multiple social media platforms has amplified the expression of public opinions over the Internet in languages such as English, Hindi and Spanish. With the aid of technological advancements in machine learning, we can analyze opinions posted on the Internet and gauge public sentiments. There are organizations and businesses that are interested in the evaluation of these sentiments as these type of data can generally be used to obtain the opinion of a product, restaurant, a candidate, etc. In this study, we perform a comparative analysis of three popular ensemble learning methodologies (Boosting, Bagging and Stacking) based …

Detecting Ai-Generated News Articles Using Unsupervised Machine Learning Algorithms, Lilou Sicard-Noel

Master's Projects

The widespread adoption of Large Language Models (LLMs) has revolutionized text generation and heightened concerns over misinformation and the erosion of journalistic integrity. Detecting AI-generated text is critical to addressing these challenges, yet current detection methods face adaptability, scalability, and accuracy limitations. This research paper uses machine-learning techniques to explore the classification of human and AI-generated articles, including a mix of human and AI-written content. The primary focus is on evaluating the effectiveness of clustering algorithms (K-Means and Agglomerative Clustering), auto-encoders, and Part-Of- Speech Tag Transition Matrix Log-Likelihood for distinguishing between AI-generated and human-written texts. Our findings reveal that while …

Image-To-Text Transcription: Analyzing And Describing Visual Content, Zixiao Fan

Master's Projects

Image captioning, which provides a textual understanding of visual content, is the fundamental support for the advancement of Human-A.I. Interaction technology. In the hope of exploring the application of such technology, this project focuses on two specific goals. One is to directly explore the application of the image informationretrieving abilities, and the other is to dive into the specifics of the pipeline and components of image captioning models. As a result, this project presents a working app that exploits the text retrieval functionalities to enable image storage with functions like tagging and transcription. It also supports search functionality with a …

Extraction Of A Knowledge Graph Of Biomedical Relationships, Brian Tran

Master's Projects

Rapid release in biomedical literature poses a challenge in linking information. This thesis aims to extract data from expanding datasets to identify and form meaningful relationships between biomedical entities. Large language models (LLMs) enable us to learn at a rapid pace. Creation of LLms from scratch are impractical. This thesis aims to collect a small dataset, containing biomedical papers, and use it to train large language models (LLMs) to extract entities from the text and learn the relationships between these entities. The experiment will be divided into two stages and utilize EU-ADR and ChemProt dataset. Starting with named entity recognition …

Galora: A Lightweight Graph-Aware Llm Framework For Node Classification On Text-Attributed Graphs, Mayur Choudhary

Master's Projects

With the exponential rise of language models (LMs) and their potential to understand semantic relationships, large LMs are being used across a wide range of applications. Text-attributed graphs (TAGs) are one notable example where LLMs can be combined with Graph Neural Networks (GNNs) to enhance node classification results. TAGs associate textual content with each node and are commonly seen in various domains such as social networks, citation graphs, recommendation systems, etc. Effectively modeling TAGs would enable deeper insights into different aspects of the graph and improve decision-making in relevant domains. We present GaLoRA, a parameter-efficient framework to integrate structural information …

Optimizing Integrated Access And Backhaul Topology Using Monte Carlo Tree Search, Srimanth Reddy Ummenthala

Master's Projects

Integrated Access and Backhaul (IAB) plays a central role in enabling scalability and high-throughput in wireless networks, especially in areas where wired backhaul is not feasible. Optimizing the topology of IAB networks is a complex task, involving trade-offs between path loss, node capacity, and link quality. To address this, this project investigates a Monte Carlo Tree Search (MCTS)-based approach to improve overall network performance by maximizing the capacity of network, considering Signal-to-Noise Ratio (SNR) on wireless links. MCTS provides a guided search mechanism to construct topologies efficiently under the constraints and is evaluated against baseline topologies. Experimental results show that …

Synthetic Malware Generation Using Generative Ai, Phanidhar Sai Sravan Chandana

Master's Projects

Malware grows in numbers and complexity, evading conventional signature-and anomaly-based defenses and worsening extreme data sparsity and class imbalance problems for machine learning based detection. Generative models, specifically GANs conditioned on contextual embeddings like BERT have proved effective augmenting training corpora to improve classifier accuracy, but these approaches have largely produced family-specific samples In this paper, we propose a generalized augmentation scheme for generating robust malware embeddings for various families. We begin by extracting opcode sequences from 13 malware families and encoding them into three embedding methods: CountVectorizer, TF-IDF, and BERT’s ‘[CLS]‘ vectors. We therefore train standard GANs and Wasserstein …

Strengthening The Preservation Ecosystem For Below-Market-Rate Housing In Santa Cruz, Bennett Williamson

Master's Projects

In Santa Cruz, California, policies and political movements of the last fifty years have both prevented sprawl and stymied the construction of new infill housing. As a result, today Santa Cruz is one of the most expensive housing markets in the nation, and the lack of available affordable housing has led to incredibly high rent burdens for the majority of the population, putting residents at risk of displacement through landlord action or market forces. This report assesses the efficacy of preserving existing affordable housing as an anti-displacement strategy in Santa Cruz. It uses key stakeholder interviews with local housing practitioners …

Nontraditional Students Enrolling In Online Education: Assessing Outcomes And Barriers To Success, Kua Vang

Master's Projects

Online education has significantly reshaped how colleges and universities define and evaluate academic achievement. In March 2020 during the height of the Covid-19 pandemic, the U.S. government enforced social distancing regulations that affected operational continuity for higher education institutions. The need to transition from traditional classroom instruction to online was unavoidable. Institutions were not prepared, however, and their shortcomings in providing high-quality online learning for students were exposed. The inadequacy can be due to higher education institutions decade-long mindset of seeing online education as a less credible alternative. As a result, institutions are more eager than ever before to invest …

Collaborative Governance In Practice: Evaluating Intergovernmental Relations Through The Santa Clara County Healthy Cities Initiative, Astrid J. Robles

Master's Projects

In the United States, intergovernmental relations (IGR) is rooted in the principles of American federalism, which focuses on the constitutional power dynamics between national, state, and local governments. While federalism outlines the structural framework, IGR specifically focuses on how federal, state, and local governments interact administratively, financially, and politically within the federal system. Beginning in the 1960’s, governments across the U.S. began to rely on non-governmental and private sector organizations for program implementation (Boyd & Fauntroy, 2000). This shift gave rise to the concept of collaborative governance, which expanded the scope of IGR by incorporating traditionally excluded groups from the …

Why We Need To Fill In Our Mixed-Use Communities’ Business Vacancies And How We Can Do That, Victor Huynh

Master's Projects

This planning report provides an in-depth exploration as to why so many mixed use communities, including transit-oriented developments, in metropolitan regions across the country are facing challenges with filling in their retail business vacancies at the ground floor. It also provides a current snapshot of best practices to remedy this widespread barrier to cities functioning more effectively and sustainably. This report provides a combination of findings from a literature review which identifies the most common reasons and remedies for business vacancies, and a research design which identifies which urban planning or urban design characteristics correlate most frequently with case sites …

Bot Detection In Social Media Using Graphsage And Bert, Abhishek Deshmukh

Master's Projects

This project details a novel bot detection system developed to battle the ever- changing challenge of disinformation, misinformation, and other bot-generated content.

The methodology employed in this project combines the text-based analytical strength of BERT (Bidirectional Encoder Representations from Transformers) with the strength of GraphSage (Graph Sample and Aggregation) for analyzing network structures. The project concatenates BERT and GraphSage vectors to create an 896-size feature embedding with a rich blend of network and text features. This project employs a Support Vector Machine to process the concatenated embeddings, as SVM works well with high-dimensional data. This project was evaluated on two …

Social Engineering Scenario Generation For Awareness-Based Attack Resilience, Jade Webb

Master's Projects

Social engineering is found in a strong majority of cyberattacks today, as it is a powerful manipulation tactic that does not require the technical skills of hacking. Calculated social engineers utilize simple communication to deceive and exploit their victims, all by capitalizing on the vulnerabilities of human nature: trust and fear. When successful, this inconspicuous technique can lead to millions of dollars in losses. Social engineering is not a one-dimensional technique; criminals often leverage a combination of strategies to craft a robust yet subtle attack. In addition, offenders are continually evolving their methods in efforts to surpass preventive measures. A …

Cca Analysis Using Computer Vision Techniques, Rahul Thakur

Master's Projects

Coral reefs are an essential part of the marine ecosystem. They perform a wide variety of tasks, some directly and others indirectly. They can produce oxygen, absorb carbon dioxide, along with supporting ocean habitat. Crustose Coralline Algae (“CCA”) plays an important role in helping provide structural support to Coral Reef ecosystems. However, global warming is causing ocean water to become more acidic resulting in coral bleaching. This is leading to changes in coral environments and causing coral deaths at alarming rates. Object detection using computer vision techniques, specifically deep learning, can help to monitor coral reef health and identify CCA …

Machine Learning Based Network Traffic Classification With Cosine-Similarity Based Out-Of-Distribution Detection, Prabhat Edupuganti

Master's Projects

The changes occurring in the amount of encrypted network traffic is growing at an alarming rate. This development has created intricate problems in traffic classification which is vital for effective cybersecurity. Moreover, most frameworks seem to ignore OOD detection, model calibration and novel pattern detection as cornerstone problem areas. The due analysis is presented as a machine learning approach aimed at resolving encrypted traffic classification issues and focuses on novel OOD detection and calibration issues. Primary contributions comprise detection of out-of-distribution states using softmax scaled cosine similarity, advanced variance-based feature elimination, and lowering ECE using stringent NNs. This work demonstrates …

Real-Time Adaptive Framework For Topic Modeling In Social Engineering Attacks, Manav Bhasin

Master's Projects

Detecting social engineering attempts is crucial for security, as these threats are becoming more frequent and increasingly exploit human vulnerabilities. This research focuses on topic modeling using conversational data from Kevin Mitnick’s ”The Art of Deception” with dialogues that illustrate various social engineering strategies. The dataset comprises manually extracted and synthetically augmented conversations to ensure natural dialogue flow. Two methodologies are presented for utterance-level and global topic extraction: prompt engineering leveraging OpenAI’s GPT-4o-mini, characterized by few-shot learning and chain-of-thought prompting, and Quantized Low Rank Adaptation (QLoRA) utilizing Mistral’s 7B instruct model for efficient fine-tuning. Through experimentation and evaluation, this study …

Separation And Surface Examination Of Spacecraft Cabin Particulates, Grace A. Belancik

Master's Theses

Particulate matter in the atmosphere is a known detriment to human health, but several factors affect any particulate’s particular toxicity including particle size, surface area, and chemical composition. Extensive studies have been conducted to examine particulate hazards on Earth, but in crewed spacecraft, like the International Space Station (ISS), the particulate environment is wholly unique due to both the controlled environment and lack of gravity. Atmospheric sampling studies are underway, but another convenient source of airborne particulates for analysis is the contents of the vacuum bags used by the astronauts to clean the air vents. However, no method currently exists …

"Acoustical Bedlam": A Contextualization Of Beethoven's Christus, Leanna R. York

The Beethoven Journal

The premiere of Beethoven’s Christus am Ölberge at the Theater an der Wien on April 5, 1803, marked the composer’s first and only contribution to the complex tradition of the German oratorio. Since the genre’s inception at the turn of the eighteenth century, it reflected artistic, literary, and theological controversies of its day. Beethoven’s chosen theme, the passion of Christ, came with its own set of ever-shifting compositional and textual expectations, influenced by older traditions of sepolcri and staged oratorios (common in Vienna and at the court of the elector of Cologne after mid-century) and lyric works such as Graun’s …

Avatars And Digital Identity: Branding In The Metaverse, Gema Bonales-Daimiel, Belén Moreno-Albarracín

Faculty Research, Scholarly, and Creative Activity

Considering avatars as virtual identities in digital universes, and according to the scarcity of previous research on the impact generated by them, this study delves into avatars’ emergence as personal brands within metaverses, assessing the mediation of AI in the design process and evaluating the impact of avatars on user experience and advertising. After carrying out a mixed methodology based on a survey to users, interviews to active companies in metaverses and avatars design specialists; and a content analysis of metaverse platforms, it’s been proved that avatars are revolutionizing digital communications, allowing immersive interactions and the implementation of virtual experiences …

Utilization Of A Single Preferred Aac System In Us Schools, Jill E. Senner, Wendy Quach, Yoosun Chung, Brian Patterson, Amy Goldman, Sarah Blackstone, Kathleen A. Post

Faculty Research, Scholarly, and Creative Activity

The use of a single preferred communication book, board, app, or device across a school or district is a frequent topic on social media. The purposes of this study were to determine: (1) When a single system is adopted by a school or district, how does frequency of input from stakeholders compare to when a single system is not used; (2) When a single system is used, are data collected; and (3) When a single system is adopted, are criteria specified and what criteria are used to determine which students receive which levels of support? A link to an online …

Grand Challenges For Human Factors And Ergonomics, Waldemar Karwowski, Gavriel Salvendy, Mica Endsley, William Rouse, Paul Salmon, Kay Stanney, Andrew Thatcher, Tony Andre, Jessie Yang, Hasan Ayaz, Ahmet Cakir, Vince Duffy, Colin Drury, Qin Gao, Yinni Guo, Peter Hancock, William S. Marras, Patrick Rau

Faculty Research, Scholarly, and Creative Activity

Contemporary society faces a growing set of complex global issues representing significant human health, well-being, and sustainability threats. Human Factors and Ergonomics (HFE) has a critical role to play in responding to these issues; however, there remain a set of grand challenges that require resolution. This paper presents and discusses six grand challenges for HFE and related key research thrust areas for each of the challenges. The grand challenges are (1) Evolution in Societal Thinking; (2) Future of Human Work in Industry 5.0; (3) Climate Change and Sustainability; (4) Future of Education and Training; (5) Future of Personalized Health, and …

Development Of Physics-Based Deterioration Models For Reinforced Soil Retaining Structures, Amr Morsy, Islam Ebo

Mineta Transportation Institute

Reinforced soil walls are key earth retention features in the transportation infrastructure. They are used to support and retain soil in a wide variety of crucial structures, such as highways, bridges, and railways, to ensure stability. They also provide solutions for constructing embankments and slopes in constrained spaces, allowing for efficient land use and improved infrastructure planning. This study used advanced numerical modeling to improve the understanding of the behavior and long-term performance of the aging reinforced soil walls from the 1970s for asset management purposes. An asset-scale model was created to simulate the effects of weather on these walls. …

Inspection Technologies For Construction And Maintenance Of Highway Infrastructure – Review And Analysis, Manideep Tummalapudi, Prem Raj Timilsena

Mineta Transportation Institute

The Federal Highway Administration (FHWA) and state Departments of Transportation (DOTs) seek to provide high quality infrastructure that meets or exceeds the desired quality standards crucial for ensuring safety and efficiency in our nation's transportation systems. Efficient inspection practices—which enable conformance with plans, specifications, and quality standards—are an essential part of this effort. In this digital age, state DOTs are relying on emerging technologies to improve inspection practices. However, there is limited knowledge available regarding various emerging technologies used for highway infrastructure construction and maintenance inspection. This study identifies various emerging technologies that are implemented at the state DOTs and …

Conceptual Critical Success Factors Model On Infrastructure Sustainability Rating System For California Construction Projects, Joseph J. Kim, Jose A. Arroyo-Turcios

Mineta Transportation Institute

This report addresses the sensitivity and reliability of the sustainability rating systems by comparing each category’s verified scores with its respective submitted scores to evaluate how points are awarded for infrastructure projects, examine which categories present the most challenges for verification, and identify the best category for verification in the sustainability rating system. The authors conducted three analyses using credit score data obtained from fourteen actual civil infrastructure projects certified under Envision. First, the Natural World category had the highest average score from the submitted and verified data. Pairwise comparisons using t-tests indicate that the mean value of one category …

Survey Of Building Information Modeling For Infrastructure (Bim4i), Maria Calahorra-Jimenez, Nigel Blampied, Elhami Nasr, Tariq Shehab

Mineta Transportation Institute

The rapid development of information technologies is transforming how data and information are produced, shared, exchanged, and managed. This transformation is accelerating in state departments of transportation (DOTs) across the country due to the pressing need for efficient means of delivering transportation projects and an enhanced need for internal and external collaboration. A key driver for this transformation is the implementation of Building Information Modeling for Infrastructures (BIM4I). The primary objective of this research was to develop actionable recommendations for DOTs to facilitate the effective adoption of BIM4I, based on national and international lessons learned and best practices. A four-step …

Understanding Mobility-Related Challenges For Aapi Older Adults: A Preliminary Study In Southern California, Yongping Zhang, Priscila Salgado Inzunza, Carol Kachadoorian, Wen Cheng, Calvin Wong

Mineta Transportation Institute

The Federal Highway Administration (FHWA) and state Departments of Transportation (DOTs) Nationwide, the Asian American Pacific Islander (AAPI) community is projected to constitute 11 percent of people 65 years and older in the United States by 2050 (He et al., 2005). The challenges limiting the transportation and mobility of AAPI older adults include, but are not limited to, language barriers, cultural barriers, anti-Asian hate, accessibility to public transit, traffic safety and public security concerns, and changes to mobility due to the COVID-19 pandemic. This project conducted an extensive literature review and a preliminary multi-language survey in Southern California to better …