Digital Commons Network^™

Efficient Multimodal Streaming Recommendation Via Expandable Side Mixture-Of-Experts, Yunke Qu, Liang Qu, Tong Chen, Quoc Viet Hung Nguyen, Hongzhi Yin

Research outputs 2022 to 2026

Streaming recommender systems (SRSs) are widely deployed in real-world applications, where user interests shift and new items arrive over time. As a result, effectively capturing users' latest preferences is challenging, as interactions reflecting recent interests are limited and new items often lack sufficient feedback. A common solution is to enrich item representations using multimodal encoders (e.g., BERT or ViT) to extract visual and textual features. However, these encoders are pretrained on general-purpose tasks: they are not tailored to user preference modeling, and they overlook the fact that user tastes toward modality-specific features such as visual styles and textual tones can …

Detecting Generative-Ai-Enabled Polymorphic Malware: A Semantic-Behavior Approach, Allyson M. Morris

Cybersecurity Undergraduate Research Showcase

AI drastically reduces the effort required to produce malware that mutates both its code and behavior, thereby creating polymorphic and nondeterministic variants in which traditional signatures and many heuristic defenses fail. In this paper, I survey recent developments in AI-assisted malware generation, explain why conventional defenses are insufficient, and propose a layered detection architecture emphasizing semantic behavior, streaming anomaly detection, and defensive generative augmentation. I also outline why this approach generalizes to previously unseen AI-mutated samples, provide an evaluation plan with meaningful metrics, and describe a feasible MVP roadmap for practical deployment. Recent disclosures and threat intelligence further highlight the …

A Systematic Review Of Poisoning Attacks Against Large Language Models (Llm), Patrick Mcguffin

Cybersecurity Undergraduate Research Showcase

This paper provides a comprehensive review of poisoning attacks against large language models (LLMs), drawing primarily from Fendley et al. (2025) and complementary studies from 2022–2025. It categorizes poisoning research into two key dimensions, Metrics and Specifications, to evaluate how attack success is measured and how attacks are implemented. This paper synthesizes quantitative results, experimental findings, and defense strategies across data, model, and multi-modal poisoning contexts. Finally, it highlights emerging challenges posed by self-adaptive and synthetic-data-driven LLMs, and proposes future research directions to strengthen model security and reliability.

Investigating The Security Vulnerabilities Of Ip Cameras: Classifications And Trends From Public Cve Data, Sam Oliver

Cybersecurity Undergraduate Research Showcase

Internet of Things (IoT) devices are increasingly targeted by cyber attacks due to weak authentication, insecure communication protocols, outdated firmware, and many other vulnerabilities. Internet Protocol (IP) cameras, a subset of these IoT devices, are particularly vulnerable and often transmit sensitive information. This paper analyzes vulnerability data from the National Vulnerability Database (NVD) to classify security vulnerabilities affecting IP cameras. Using this dataset, the paper examines the types and frequencies of these vulnerabilities, including authentication bypass, web interface exploits, and default and weak credentials. We additionally examine trends over time and across categories. This research aims to identify the primary …

Software-Defined Networking Powered By Ai-Driven Anomaly Detection, Dina Moloja, Vusumuzi Malele Prof

Journal of Cybersecurity Education, Research and Practice

Software Defined Networking (SDN) revolutionizes network control by separating the control plane from the data plane. Although the latter improves SDN agility and scalability, it creates a security hole, particularly in a central control plane, leading to SDN environments becoming high-profile targets for advanced cybersecurity threats. Due to static and signature-based point-in-time behavior, traditional security methods are unable to keep up with modern attacks that are an anomaly to SDNs. Artificial Intelligence (AI) with its different applications and techniques, has the capability of detecting SDN cyber threats’ anomalies. This paper presents the results of a literature scoping exercise that used …

Multi-Resolution Graph Neural Networks For Spread Prediction, Petr Kisselev

Annual Symposium on Biomathematics and Ecology Education and Research

No abstract provided.

Evaluation Of Integrating Smart Technologies Into Business Ecosystems Using Neutrosophic Uncertainty Model, Kainat Muniba, Muhammad Naveed Jafar, Adil Ahmad, Aruna Pavate

Neutrosophic Systems with Applications

This study proposes a methodological framework for Evaluation of Integrating Smart Technologies into Business Ecosystems. We use the decision-making process to deal with different criteria and alternatives. The decision-making process is used under the neutrosophic set to overcome uncertainty information. Neutrosophic set has three membership functions such as truth, indeterminacy, and falsity. These functions are used to overcome vague information. The average method is used to compute the criteria weights. The COBRA method is used to rank the alternatives based on different alternatives. This study uses 8 criteria and 15 alternatives to be evaluated to show the best option.

Neutrosophic Graded Jordan–Bialgebra Framework For Ai-Driven Analysis, Mona Gharib, Rana Muhammad Zulqarnain, Xiao Long Xin, Muhammad Gulistan, Muhammad Abid

Neutrosophic Systems with Applications

Modern Artificial Intelligence (AI) systems face significant challenges in processing and analyzing datasets characterized by high degrees of uncertainty, ambiguity, and indeterminacy, which are prevalent features in complex real-world scenarios. To address this limitation, this study introduces a novel neutrosophic graded Jordan–bialgebra framework. This framework strategically integrates the inherent structural properties of Jordan–Bialgebras with the advanced capability of Neutrosophic Graded Structures to simultaneously model degrees of truth, indeterminacy, and falsehood. The primary objective of this study is to establish a rigorous algebraic foundation that enables AI models to perform a more robust and comprehensive analysis of data containing incomplete or …

Comprehensive Risk Evaluation Framework For Reducing Errors In Critical Airport Systems, Arshad Hameed, Muhammad Umer Farooq

Neutrosophic Systems with Applications

Airports are intricate systems that are subject to a number of operational hazards that might impair safety and cause service interruptions. There is a dearth of thorough approaches designed specifically for airport operations, despite a wealth of research on risk management in several industries. In order to anticipate and prevent breakdowns in vital airport systems, we offer a thorough risk assessment approach that combines quantitative evaluation and human factor analysis. This study uses the decision-making methodology such as MARCOS method to rank the alternatives. The criteria weights are computed in this study. This study uses six criteria and 14 alternatives.

Integrated Neutrosophic Set For Assessing The Route Options In Overweight Complex Transportation Planning Scenarios, Muhammad Abid, Tayyaba Akhtar, Harshit Bhatt

Neutrosophic Systems with Applications

Due to improvements in lifting and transportation technology that enable the long-distance movement of heavy loads, overweight and oversized transport (O&OT) has emerged as one of the most important aspects of project logistics. This kind of transportation operation, also known as abnormal transportation, is heavily influenced by external variables like weather, traffic density, and legal regulations, as well as technical factors like the load’s weight and geometry, road surface, axle load limitations, slope, and ground strength. Decision-Makers (DMs) and practitioners who plan and carry out operations without giving these variables and factors enough thought may cause operational delays, significant hazards, …

Evaluating Solar Energy Challenges And Strategies To Overcome Challenges Under Neutrosophic Decision Making Methodology, Ahmed A El-Douh, Mina Samir Shenouda

Neutrosophic Systems with Applications

Because of its reliance on fossil fuels and fast population expansion, the nation confronts severe socioeconomic problems that can result in political instability, environmental damage, health issues, and economic instability. A switch to clean energy is now necessary as a result of these problems widening the gap between supply and demand for energy. To close this gap, photovoltaic (PV) technology holds great potential. In order to reduce bias and handle ambiguity while assessing solar energy (SE) issues and regulations, this study offers a decision-making process in a single valued neutrosophic set (SVNS) environment. The neutrosophic set is used to overcome …

Sound Practices In Human Resources: A Critical Analysis Of How Ai Applicant Screening Tools Are Used With Deaf Individuals In The Hospitality And Tourism Sector, Sherry L. Gaynor

USF Tampa Graduate Theses and Dissertations

This study explores the intersection of Career and Workforce Education, Deaf Education, and artificial intelligence (AI) technologies used in hiring practices. The purpose of this research is to identify potential barriers to equitable employment for Deaf and Hard-of-Hearing (DHH) individuals and to propose effective methods for evaluating and assessing AI and machine learning applications utilized in recruitment and selection processes. Although AI is increasingly embedded in human resource management (HRM), there remains a significant lack of research addressing its use with special populations. In particular, no prior studies have examined how voice recognition software, facial analysis, psychometric testing, and virtual …

From The Editors, Rully Karim Dr.

Journal of Project Management & Construction

The Journal of Project Management and Construction (JPMC) is a peer-reviewed publication dedicated to advancing the field of project management and construction, grounded in the principles outlined in the PMBOK 6th Edition. Our focus encompasses the ten knowledge areas essential to successful project management: Integration, Scope, Schedule, Cost, Quality, Resource, Communications, Risk, Procurement, and Stakeholder Management.

JPMC publishes original research papers written in English that provide deeper insights into these knowledge areas and contribute to the development of best practices in project management and construction. Submissions may include theoretical analyses, computational models, experimental observations, or a combination of both theoretical …

Computational Data Analysis, Kathryn S. Biles

LSU Master's Theses

Data science has emerged as a cornerstone of innovation, shaping an ever-expanding range

of professional careers. As technology advances and the volume of data expands expo-

nentially, the ability to extract meaningful insights from data has become indispensable

across industries. Far from representing a single career path, data science enables profes-

sionals in nearly every domain to make informed decisions, optimize systems, and drive

innovation. Yet, many high school students and incoming college freshmen have limited

exposure to data science fundamentals or the career opportunities they unlock. This is

the gap that Computational Data Analysis, a high school-level curriculum I …

Preparing Tomorrow’S Professionals: Industry-Informed Ai Integration, Brent A. Terwilliger Ph.D, John Faraca

Publications

As AI reshapes operations across aviation and aerospace, organizations are investing in ways to preserve data integrity, safeguard proprietary knowledge, and uphold critical professional competencies. This presentation shares emerging findings from a study that surveys and interviews industry professionals about their use of AI tools, their concerns about misuse, and the importance of secure, enterprise-controlled “walled garden” environments. The work explores how employers define appropriate, effective, and innovative AI adoption, particularly in roles requiring high-stakes decision-making, compliance, and technical acumen.

By analyzing organizational expectations around AI-related knowledge, skills, and abilities (KSAs), this research offers practical guidance for academic programs seeking …

Student Perspectives On Ai-Enabled Tools For Adaptive Learning, John Faraca

Publications

Artificial Intelligence (AI) is increasingly influencing the delivery of higher education, especially in aviation technical disciplines. From AI-assisted gimbals and video production tools to generative AI platforms, these technologies are helping learners to engage with course material, accomplish objectives, and connect academic concepts with professional applications. By offering pathways for personalization, streamlining resource access, and supporting interactive instruction, AI tools expand opportunities for effective learning. This work builds on a current collaborative research project with a faculty researcher to explore the student perspective in the active review and application of these tools to highlight their potential to improve usability, address …

Advancing Power System Reliability And Security With Efficient And Resilient Graph Neural Network Frameworks, Seyed Hamed Haghshenas

USF Tampa Graduate Theses and Dissertations

Enhancing the reliability and security of smart grids is critical for ensuring their seamless operation and resilience against disruptions. The increasing integration of distributed energy resources, advanced measurement devices, and cyber-physical elements introduces both opportunities and challenges for grid management. While these advancements provide enhanced visibility and operational control, they also expose the grid to vulnerabilities from cyber-physical stresses, such as cyber-attacks, equipment failures, and fluctuating power demands. Traditional methods for reliability assessment and threat detection often rely on model-based approaches that struggle to adapt to the complexity and dynamic nature of modern smart grids. These limitations necessitate novel data-driven …

Effects Of Code Scaffolding In Increasing Student Confidence In Programming Cryptography, John Denny

LSU Master's Theses

Cryptography is essential for secure communications, and new threats require more students willing to program and interact with cryptographic systems. Previous research is focused on tools for teaching these systems at a high level, teaching through attacks against these systems, and proper use of these systems in software development. In this paper, we seek to design a workshop to use scaffolded Python code to teach how these cryp- tographic systems are designed. We explore the use of code scaffolding for students to program an example implementation of the McEliece crypto- graphic system to build confidence in working with these systems. …

Systematic Review Of Elementary Cybersecurity Education: Curriculum, Pedagogy, And Barriers, Na Liu, Siyu Long, Florence Martin

Journal of Cybersecurity Education, Research and Practice

Abstract -As children increasingly engage with digital platforms, the need for effective cybersecurity education has become urgent. This systematic review synthesizes 81 studies published between 2017 and 2024 to examine global curricula research focus and topics, pedagogical approaches and assessment methods, and key challenges in elementary cybersecurity education. The findings reveal six major thematic categories: student awareness, parental mediation, teacher engagement, curriculum design, community and policy support, and pedagogical innovation. Among instructional strategies, game-based learning and narrative storytelling emerge as the most frequently explored. Despite this growth, major gaps remain in curriculum consistency, teacher preparation, assessment rigor, and stakeholder coordination. …

Experts’ Validation Of The Fundamental Cybersecurity Competency Index (Fcci) Using A Commercial Cyber Range Through Human-Generative Artificial Intelligence (Genai) Teaming, Dariusz Witko, Yair Levy, Catherine Neubauer, Greg Simco, Laurie P. Dringus, Melissa Carlton

Journal of Cybersecurity Education, Research and Practice

The increasing volume of cyber threats, combined with a critical shortage of skilled professionals and rising burnout among practitioners, highlights the urgent need for innovative solutions in cybersecurity operations. Generative Artificial Intelligence (GenAI) offers promising potential to augment human analysts in cybersecurity, but its integration requires rigorous validation of the fundamental competencies that enable effective collaboration of human-GenAI teams. This research study employed a mixed-methods research project designed to evaluate human-GenAI teams, emphasizing the role of expert consensus in shaping the experimental assessment of the Fundamental Cybersecurity Competency Index (FCCI) in a commercial cyber range. We engaged 20 Subject Matter …

Exploring Runtime Evolution In Android: A Cross-Version Analysis And Its Implications For Memory Forensics., Babangida Bappah

LSU Master's Theses

Userland memory forensics has become a critical component of smartphone investigations and incident response, enabling the recovery of volatile evidence such as deleted messages from end-to-end encrypted apps and cryptocurrency transactions. However, these forensics tools, particularly on Android, face significant challenges in adapting to different versions and maintaining reliability over time due to the constant evolution of low-level structures critical for evidence recovery and reconstruction. Structural changes, ranging from simple offset modifications to complete architectural redesigns, pose substantial maintenance and adaptability issues for forensic tools that rely on precise structure interpretation. Thus, this paper presents the first systematic study of …

Enhancing Cloud-Based Threat Detection Through Explainable Ai: A Comparative Study Of Machine Learning And Xai-Integrated Models, Amna Al Ghaithi

Thesis/ Dissertation Defenses

The rapid adoption of cloud computing brought about serious security concerns, as cloud infrastructures are constantly exposed to cybersecurity threats such as malware and Distributed Denial of Service attacks. Also, on the other hand, current security methodologies have limitations in identifying new threats accurately. Apart from the fact that ML models are highly efficient in detecting attacks, as ‘black boxes,’ they lack interpretability, impacting trust and adoption within vital cloud environments. This research aims to solve this issue by integrating Explainable Artificial Intelligence practices to help enhance both the accuracy and interpretability of AI systems intended to detect threats in …

Seeing Culture: A Benchmark For Visual Reasoning And Grounding, Burak Satar, Zhixin Ma, Patrick Amadeus Irrawan, Wilfried Ariel Mulyawan, Jing Jiang, Ee-Peng Lim, Chong-Wah Ngo

Research Collection School Of Computing and Information Systems

Multimodal vision-language models (VLMs) have made substantial progress in various tasks that require a combined understanding of visual and textual content, particularly in cultural understanding tasks, with the emergence of new cultural datasets. However, these datasets frequently fall short of providing cultural reasoning while underrepresenting many cultures.In this paper, we introduce the Seeing Culture Benchmark (SCB), focusing on cultural reasoning with a novel approach that requires VLMs to reason on culturally rich images in two stages: i) selecting the correct visual option with multiple-choice visual question answering (VQA), and ii) segmenting the relevant cultural artifact as evidence of reasoning. Visual …

Adasteer: Your Aligned Llm Is Inherently An Adaptive Jailbreak Defender, Weixiang Zhao, Jiahe Guo, Yulin Hu, Yang Deng, An Zhang, Xingyu Sui, Xinyang Han, Yanyan Zhao, Bing Qin, Tat-Seng Chua, Ting Liu

Research Collection School Of Computing and Information Systems

Despite extensive efforts in safety alignment, large language models (LLMs) remain vulnerable to jailbreak attacks. Activation steering offers a training-free defense method but relies on fixed steering coefficients, resulting in suboptimal protection and increased false rejections of benign inputs. To address this, we propose AdaSteer, an adaptive activation steering method that dynamically adjusts model behavior based on input characteristics. We identify two key properties: Rejection Law (R-Law), which shows that stronger steering is needed for jailbreak inputs opposing the rejection direction, and Harmfulness Law (H-Law), which differentiates adversarial and benign inputs. AdaSteer steers input representations along both the Rejection Direction …

Chain Of Strategy Optimization Makes Large Language Models Better Emotional Supporter, Weixiang Zhao, Xingyu Sui, Xinyang Han, Yang Deng, Yulin Hu, Jiahe Guo, Libo Qin, Qianyun Du, Shijin Wang, Yanyan Zhao, Bing Qin, Ting Liu

Research Collection School Of Computing and Information Systems

The growing emotional stress in modern society has increased the demand for Emotional Support Conversations (ESC). While Large Language Models (LLMs) show promise for ESC, they face two key challenges: (1) low strategy selection accuracy, and (2) preference bias, limiting their adaptability to users’ emotional needs. Existing supervised fine-tuning (SFT) struggles to address these issues, as it rigidly trains models on single gold-standard responses without modeling nuanced strategy trade-offs. To overcome these limitations, we propose a novel two-stage framework that optimizes strategy selection preferences at each dialogue turn. We first leverage Monte Carlo Tree Search to construct ESC-Pro, a high-quality …

Energy-Harvesting Concurrent Lora Mesh With Timing Offsets For Underground Mine Emergency Communications, Hilary Kelechi Anabi, Samuel Frimpong, Sanjay Madria

Mining Engineering Faculty Research & Creative Works

Underground mine emergencies destroy communication infrastructure when situational awareness is most critical. Current systems rely on centralized network infrastructure, which fails during emergencies when miners are trapped and require rescue coordination. This paper proposes an energy-harvesting LoRa mesh network that addresses self-powered operation, interference management, and adaptive physical layer optimization under severe underground propagation conditions. A dual-antenna architecture separates RF energy harvesting (860 MHz) from LoRa communication (915 MHz), enabling continuous operation with supercapacitor storage. The core contribution is a decentralized scheduler that derives optimal timing offsets by modeling concurrent transmissions as a Poisson collision process, exploiting LoRa's capture effect …

Composition Pedagogy As Ai‑Native Coding: From Design Kit To Scholarly Framework, Daniel Plate, James Hutson

Faculty Scholarship

This article advances a field-ready framework that reconceives first-year composition as AI-native coding, translating a complete “design kit” into scholarly method, evaluative protocol, and curriculum architecture. Background: Contemporary composition pedagogy emphasizes process, genre awareness, and collaborative revision; meanwhile, modern software practice operationalizes iteration through version control, test-driven development, and continuous integration. The uploaded kit demonstrates that these cultures are isomorphic: writing stages align with SDLC phases, and automated pipelines can lint prose, execute argument “tests,” and publish artifacts with auditable histories. Approach: The study systematizes that kit into (1) a conceptual map that recasts authorship as orchestration and verification, (2) …

Navigating Equity In The Digital Era : Addressing Challenges And Advancing Rights Of Female Seafarers Through Policy Reform, And Artificial Intelligence, Margaret Dixon

World Maritime University Dissertations

No abstract provided.

When Deep Learning Meets Information Retrieval-Based Bug Localization: A Survey, Feifei Niu, Chuanyi Li, Kui Liu, Xin Xia, David Lo

Research Collection School Of Computing and Information Systems

Bug localization is a crucial aspect of software maintenance, running through the entire software lifecycle. Information retrieval-based bug localization (IRBL) identifies buggy code based on bug reports, expediting the bug resolution process for developers. Recent years have witnessed significant achievements in IRBL, propelled by the widespread adoption of deep learning (DL). To provide a comprehensive overview of the current state of the art and delve into key issues, we conduct a survey encompassing 61 IRBL studies leveraging DL. We summarize best practices in each phase of the IRBL workflow, undertake a meta-analysis of prior studies, and suggest future research directions. …

Do Code Semantics Help? A Comprehensive Study On Execution Trace-Based Information For Code Large Language Models, Jian Wang, Xiaofei Xie, Qiang Hu, Shangqing Liu, Yi Li

Research Collection School Of Computing and Information Systems

Code Large Language Models (Code LLMs) have opened a new era in programming with their impressive capabilities. However, recent research has revealed critical limitations in their ability to reason about runtime behavior and understand the actual functionality of programs, which poses significant challenges for their post-training and practical deployment. Specifically, Code LLMs encounter two principal issues: (1) a lack of proficiency in reasoning about program execution behavior, as they struggle to interpret what programs actually do during runtime, and (2) inconsistent and fragmented representation of semantic information, such as execution traces, across existing methods, which hinders their ability to generalize …