Digital Commons Network^™

From Philosophy To Nlu: Evolving Definitions Of Research Hypotheses, Jian Wu, Sarah Rajtmajer

Computer Science Faculty Publications

Over the past decades, alongside advancements in natural language processing, significant attention has been paid to training models to automatically extract, understand, test, and generate hypotheses in open and scientific domains. However, interpretations of the term hypothesis for various natural language understanding (NLU) tasks have migrated from traditional definitions in the natural, social, and formal sciences. Even within NLU, we observe differences defining hypotheses across literature. In this paper, we overview and delineate various definitions of hypothesis. Especially, we discern the nuances of definitions across recently published NLU tasks. We highlight the importance of well-structured and well-defined hypotheses, particularly as …

S²Il: Structurally Stable Incremental Learning, S. Balasubramanian, P. Yedu Krishna, Talasu Sai Sriram, M. Sai Subramaniam, Manepalli Pranav Phanindra Sai, Ravi Mukkamala

Computer Science Faculty Publications

Feature Distillation (FD) strategies are proven to be effective in mitigating Catastrophic Forgetting (CF) seen in Class Incremental Learning (CIL). However, current FD approaches enforce strict alignment of feature magnitudes and directions across incremental steps, limiting the model’s ability to adapt to new knowledge. In this paper, we propose Structurally Stable Incremental Learning (S²IL), a FD method for CIL that mitigates forgetting by focusing on preserving the overall spatial patterns of features which promote flexible (plasticity) yet stable representations that preserve old knowledge (stability). We also demonstrate that our proposed method S²IL achieves strong incremental accuracy and outperforms other FD …

Understanding Pii Leakage In Large Language Models: A Systematic Survey, Shuai Cheng, Zhao Li, Shu Meng, Mengxia Ren, Haitao Xu, Shuai Hao, Chuan Yue, Fang Zhang

Computer Science Faculty Publications

Large Language Models (LLMs) have demonstrated exceptional success across a variety of tasks, particularly in natural language processing, leading to their growing integration into numerous facets of daily life. However, this widespread deployment has raised substantial privacy concerns, especially regarding personally identifiable information (PII), which can be directly associated with specific individuals. The leakage of such information presents significant real-world privacy threats. In this paper, we conduct a systematic investigation into existing research on PII leakage in LLMs, encompassing commonly utilized PII datasets, evaluation metrics, and current studies on both PII leakage attacks and defensive strategies. Finally, we identify unresolved …

Humans Vs. Llms On Open Domain Scientific Claim Verification: A Baseline Study, Benjamin Curtis, Stefania Dzhaman, Matthew Maisonave, Jian Wu

Computer Science Faculty Publications

Verifying scientific claims is challenging for the general public because most people lack domain knowledge. Manual verification by subject domain experts is accurate, but it is obviously not scalable to meet the rising number of scientific claims on the Web. Whether the emerging large language models and large reasoning models can be used for scientific claim verification, and how their performances compare to humans, are still research questions. To this end, we developed a new benchmark MSVEC2 that consists of 138 claims from credible fact verification websites and science news outlets. Two tasks were given to both human and LLM …

Privshap: A Finer-Granularity Network Linearization Method For Private Inference, Xiangrui Xu, Zhenzhen Wang, Rui Ning, Chunsheng Xiu, Hongyi Wu

Computer Science Faculty Publications

Private inference applies cryptographic techniques like homomorphic encryption, garble circuit and secret sharing to keep both sides privacy in a client-server setting during inference. It is often hindered by the high communication overheads, especially at non-linear activation layers such as ReLU. Hence ReLU pruning has been widely recognized as an efficient way to accelerate private inference. Existing approaches to ReLU pruning typically rely on coarse hypothesis, which assume an inverse correlation between the importance of ReLU and linear layers or shallow activation layers have less importance for universal models, to assign the budgets according to the layer while preserving the …

An Analytical Review Of Preprocessing Techniques In Bengali Natural Language Processing, Sovon Chakraborty, Protiva Das, Shakib Mahmud Dipto, Md Aktaruzzaman Pramanik, Jannatun Noor

Computer Science Faculty Publications

Research in Bengali Natural Language Processing (BNLP) is rapidly expanding. Despite being one of the most widely spoken languages in the world, BNLP research remains insufficient, particularly in Bengali speech recognition. The languages rich morphology, agglutinative structure, and diverse dialects make text and speech processing especially challenging. However, these challenges can be addressed with effective preprocessing techniques. Various organizations in Bangladesh and West Bengal are integrating Natural Language Processing (NLP) into their services, but without a thorough understanding of preprocessing, these implementations remain incomplete. Applying proper preprocessing techniques to the Bengali language will serve as a foundation for developing robust …

Not Here, Go There: Analyzing Redirection Patterns On The Web, Kritika Garg, Sawood Alam, Dietrich Ayala, Michele C. Weigle, Michael L. Nelson

Computer Science Faculty Publications

URI redirections are integral to web management, supporting structural changes, SEO optimization, and security. However, their complexities affect usability, SEO performance, and digital preservation. This study analyzed 11 million unique redirecting URIs, following redirections up to 10 hops per URI, to uncover patterns and implications of redirection practices. Our findings revealed that 50% of the URIs terminated successfully, while 50% resulted in errors, including 0.06% exceeding 10 hops. Canonical redirects, such as HTTP to HTTPS transitions, were prevalent, reflecting adherence to SEO best practices. Non-canonical redirects, often involving domain or path changes, highlighted significant web migrations, rebranding, and security risks. …

Human Perception Of Ai Capabilities At Classifying Perturbed Roadway Signs, Katherine R. Garcia, Jing Chen, Yanru Xiao, Scott Mishler, Cong Wang, Bin Hu

Computer Science Faculty Publications

Artificial Intelligence (AI) is crucial to numerous functions required for driving automation systems, including the computer vision techniques used to detect the roadway environment and make real-time decisions. However, the images used as inputs to the AI system may be maliciously perturbed, or manipulated, causing the AI system to make an incorrect classification. In this study, we examined humans’ perception of the AI’s computer vision capability of classifying various road sign images, including the original images, images with two different types of malicious attacks, and images that are scrambled randomly at the pixel level. Our results showed that participants rated …

A Real-Time Approach To Capture Ambient And Focal Attention In Visual Search, Gavindya Jayawardena, Yasith Jayawardana, Yasasi Abeysinghe, Bhanuka Mahanama, Sampath Jayarathna, Jacek Gwizdka

Computer Science Faculty Publications

During visual search, individuals’ attention shifts between ambient and focal states in response to task demands and stimuli. The ambient/focal coefficient K is a statistically validated measure of these states, computed offline from fixation duration and saccade amplitude data. While current methods compute K offline, real-time computation could enable applications such as monitoring user attention, creating attention-adaptive user interfaces, and optimizing graphics rendering. However, real-time computation of K requires stable estimates for the parameters of fixation duration and saccade amplitude distributions. Since these distributions are heavy-tailed, the real-time estimates exhibit high variance and slow convergence. To overcome this, we propose …

Multi-Modal Mri Based Segmentation Of Brain Metastases Using Adaptive Self-Attention, Evan Savaria

Computer Science Faculty Publications

Brain metastases (BMs) are the most common adult central nervous system malignancy, affecting 20–40% of cancer patients. Accurate segmentation of metastatic lesions in multi-modal MRI is essential for treatment planning and prognosis however, manual delineation is time consuming and prone to variability. Traditional deep learning models such as U-Net, have improved segmentation accuracy but capture limited long-range dependencies and struggle with variations in metastasis size, shape, and distribution. This study introduces the Adaptive Integrated Multi-modal Segmentation (AIMS) model, an adaptive self-attention framework within a hybrid U-Net and Transformer architecture to enhance BM segmentation by leveraging multi-modal MRI integration. The proposed …

Ai For Nuclear Physics: The Exclaim Project, S. Liuti, D. Adams, M. Boër, G. W. Chern, M. Cuic, M. Engelhardt, G. R. Goldstein, B. Kriesten, Y. Li, H. W. Lin, M. Sievert, D. Sivers

Computer Science Faculty Publications

An overview of the recent activity of the newly funded EXCLusives with AI and Machine learning (EXCLAIM) collaboration is presented. The main goal of the collaboration is to develop a framework to implement AI and machine learning techniques in problems emerging from the phenomenology of high energy exclusive scattering processes from nucleons and nuclei, maximizing the information that can be extracted from various sets of experimental data, while implementing theoretical constraints from lattice QCD. A specific perspective embraced by EXCLAIM is to use the methods of theoretical physics to understand the working of ML, beyond its standardized applications to physics …

A Survey On Deep Learning For Drug-Target Binding Prediction: Models, Benchmarks, Evaluation, And Case Studies, Kusal Debnath, Pratip Rana, Preetam Ghosh

Computer Science Faculty Publications

Conventional drug discovery is expensive, time-consuming, and prone to failure. Artificial intelligence has become a potent substitute over the last decade, providing strong answers to challenging biological issues in this field. Among these difficulties, drug-target binding (DTB) is a key component of drug discovery techniques. In this context, drug-target affinity and drug–target interaction are complementary and essential frameworks that work together to improve our comprehension of DTB dynamics. In this work, we thoroughly analyze the most recent deep learning models, popular benchmark datasets, and assessment metrics for DTB prediction. We look at the paradigm shift in the development of drug …

Accessmenu: Enhancing Usability Of Online Restaurant Menus For Screen Reader Users, Nithiya Venkatraman, Akshay Kolgar Nayak, Suyog Dahal, Yash Prakash, Hae-Na Lee, Vikas Ashok

Computer Science Faculty Publications

Online food ordering has become commonplace due to its convenience. The wide variety of culinary choices, combined with fast and economical door-delivery services, encourages more people to order food online. To facilitate this process, food vendors, including restaurants, often provide full menus on their websites, typically in visual formats such as images or PDFs. While this is convenient for sighted users, blind and visually impaired (BVI) individuals face significant challenges accessing these visual menus with their screen reader assistive technology. An interview study with 12 BVI screen reader users revealed that present assistive tools do not adequately satisfy the needs …

Deepssetracer 2.0: Improved Deep Learning Model Performance For Protein Secondary Structure Segmentation From Cryo-Em Maps, Bryan Hawickhorst, Thu Nguyen, Willy Wriggers, Jiangwen Sun, Jing He

Computer Science Faculty Publications

DeepSSETracer is a method for segmenting protein secondary structure from medium-resolution (5-10Å) cryogenic electron microscopy (cryo-EM) density maps. We conducted experiments and ablation studies to examine the effects of normalization methods, max-pooling, activation functions, and loss calculation region on DeepSSETracer. By combining multiple technical improvements, the performance of the new version, DeepSSETracer 2.0, was significantly enhanced compared to DeepSSETracer 1.1. On a set of 77 test cases, the weighted average per-voxel F1 score increased from 62.1% to 70.3% for helix detection, and from 47.8% to 62.5% for β-sheet detection. While each of the five modifications in the network enhanced the …

Benchmarking And Improving Foundation Model Dietary Estimates From Meal Images, Yongcheng Mu, Jiangwen Sun, Jing He

Computer Science Faculty Publications

Accurate quantifying dietary contents, such as calories, proteins, carbohydrates, and fats, from an image of a meal plate is vital for managing diabetes. Recently, Large Multimodal Models (LMMs) have excelled in complex vision-language tasks due to their use of very large, highly diverse data. This study benchmarked the use of seven LMMs that include full and lightweight models of GPT, Gemini, and Llama for nutrition estimation based on Google's Nutrition5k dataset and our own phone-collected DonateAndLearn dataset. We analyzed the performance of LMMs and the RGB-D fusion model, in which the RGB-D model was specifically trained using Nutrition5k data. On …

Effective Pii Extraction From Llms Through Augmented Few-Shot Learning, Shuai Cheng, Shu Meng, Haitao Xu, Haoran Zhang, Shuai Hao, Chuan Yue, Wenrui Ma, Meng Han, Fang Zhang, Zhao Li

Computer Science Faculty Publications

Large Language Models (LLMs) exhibit strong natural language processing capabilities but also pose significant privacy risks, particularly regarding the leakage of Personally Identifiable Information (PII) embedded in their training data. Existing PII extraction methods suffer from the limitations of low success rates or impracticality for large-scale PII extraction. In this study, we propose a novel PII extraction approach based on enhanced few-shot learning techniques, which achieves efficient and cost-effective PII retrieval without relying on fine-tuning or jailbreaking. We evaluated our approach on both open-source and closed-source LLMs. The experimental results demonstrate that, for non-targeted PII extraction, the attack success rate …

An Optimized Generalized Multi-Color Point Implicit Solver For Intel Gpus Using Oneapi Esimd, Joseph Wassell, Mohammad Zubair, Aaron Walden, Gabriel Nastac, Eric Nielsen, Timothée Ewart

Computer Science Faculty Publications

This paper presents an efficient implementation of a linear-solver kernel relevant to FUN3D, a suite of computational fluid dynamics software developed at NASA’s Langley Research Center. The linear solver is optimized for a range of block sizes commonly used in FUN3D. The implementation targets Aurora, the Argonne Leadership Computing Facility’s (ALCF) exascale machine featuring Intel Data Center Max 1550 GPUs. The linear solver’s performance is memory bandwidth-bound due to its low arithmetic intensity. The primary performance challenges stem from variable matrix row lengths and indirect memory access patterns inherent in unstructured-grid applications. Variable block sizes introduce additional complexity through differing …

Characterizing Language Use In Online Accessibility Discussion Forums, Nithiya Venkatraman, Anand Ravi Aiyer, Yash Prakash, Sampath Jayarathna, Hae-Na Lee, Vikas Ashok

Computer Science Faculty Publications

Discussion forums are one of the favored platforms for knowledge sharing. Given their popularity, copious research exists on understanding the linguistic and behavioral characteristics of forum conversations, so as to inform the design of many downstream applications including discourse visualization, sentiment analysis, and question answering. However, prior investigations have mainly focused on general forums designed primarily for sighted users, and as such the applicability of their findings to dedicated accessibility discussion forums frequented by blind screen reader users remains unanswered. To bridge this knowledge gap and facilitate the development of better-informed assistive technologies for blind people, we investigated language use …

Contextual Memory Recall: A Novel Metric For Class Incremental Learning, Balasubramanian S, Sai Subramaniam M., Sai Sriram Talasu, Yedu Krishna P., Pranav Phanindra Sai M., Darshan Gera, Ravi Mukkamala

Computer Science Faculty Publications

We propose a novel metric for class incremental learning (CIL) called Contextual Memory Recall (CMR), which evaluates how well a CIL model recalls previously learned classes when given relevant past cues. Inspired by human memory, CMR offers newer insights into continual aspects of a CIL model that were not addressed by previously proposed metrics for CIL. Specifically, the standard metric, average incremental accuracy (AIA), overlooks the quality of evolving feature representations, whereas our proposed CMR accounts for it. As a result, methods using feature distillation perform well under AIA but poorly under CMR, while those without feature distillation excel under …

Machine Learning Predicting Acute Pain And Opioid Dose In Radiation Treated Oropharyngeal Cancer Patients, Vivian Salama, Laia Humbert-Vidan, Brandon Godinich, Kareem A Wahid, Dina M Elhabashy, Mohamed A Naser, Renjie He, Abdallah S R Mohamed, Ariana J Sahli, Katherine A Hutcheson, Gary Brandon Gunn, David I Rosenthal, Clifton D Fuller, Amy C Moreno

Faculty, Staff and Student Publications

Introduction: Acute pain is common among oral cavity/oropharyngeal cancer (OCC/OPC) patients undergoing radiation therapy (RT). This study aimed to predict acute pain severity and opioid doses during RT using machine learning (ML), facilitating risk-stratification models for clinical trials.

Methods: A retrospective study examined 900 OCC/OPC patients treated with RT during 2017-2023. Pain intensity was assessed using NRS (0-none, 10-worst) and total opioid doses were calculated using morphine equivalent daily dose (MEDD) conversion factors. Analgesics efficacy was assessed using combined pain intensity and total MEDD. ML predictive models were developed and validated, including Logistic Regression (LR), Support Vector Machine (SVM), Random …

Artificial Intelligence In Science And Society: The Vision Of Usern, Tommaso Dorigo, Gary D. Brown, Carlo Casonato, Artemi Cerda, Joseph Ciarrochi, Mauro Da Lio, Nicole D'Souza, Nicolas R. Gauger, Steven C. Hayes, Stefan G. Hofmann, Robert Johansson, Marcus Liwicki, Fabien Lotte, Juan J. Nieto, Giulia Olivato, Peter Parnes, George Perry, Alice Plebe, Idupulapati M. Rao, Nima Rezaei, Fredrik Sandin, Andrey Ustyuzhanin, Giorgio Vallortigara, Pietro Vischia, Niloufar Yazdanpanah

All Peer-Reviewed Publications

The recent rise in relevance and diffusion of Artificial Intelligence (AI)-based systems and the increasing number and power of applications of AI methods invites a profound reflection on the impact of these innovative systems on scientific research and society at large. The Universal Scientific Education and Research Network (USERN), an organization that promotes initiatives to support interdisciplinary science and education across borders and actively works to improve science policy, collects here the vision of its Advisory Board members, together with a selection of AI experts, to summarize how we see developments in this exciting technology impacting science and society in …

A Review On Knowledge And Information Extraction From Pdf Documents And Storage Approaches, Salvador D. Atagong, Henri Tonnang, Kennedy Senagi, Mark Wamalwa, Komi M. Agboka, John Odindi

All Peer-Reviewed Publications

Introduction: Automating the extraction of information from Portable Document Format (PDF) documents represents a major advancement in information extraction, with applications in various domains such as healthcare, law, or biochemistry. However, existing solutions face challenges related to accuracy, domain adaptability, and implementation complexity. Methods: A systematic review of the literature was conducted using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) methodology to examine approaches and trends in PDF information extraction and storage approaches. Results: The review revealed three dominant methodological categories: rule-based systems, statistical learning models, and neural network-based approaches. Key limitations include the rigidity of rule-based …

Evaluating Multimodal Ai Systems: A Comparative Analysis Of Large Languagel Model-Based Models For Text, Image, And Video Generation, Azeezat O. Akinola

College of Graduate Studies: Theses & Dissertations

In the era of rapid technological advancement, efficient content generation, application development, and data management are crucial for meeting the demands of dynamic digital environments. This thesis uses state-of-the-art models to explore three core areas: AI-driven video content creation, text-to-image-to-text consistency, and automatic text summarization. The first study investigates the potential of AI-powered text-to-video generation to democratize video production and enhance storytelling. By comparing the performance of three models—ModelScope, Text2Video (Zero), and Motion Consistency—this study assessed the quality of generated videos using CLIP scores. It evaluated statistical significance through t-tests and homogeneity tests. Results indicate that ModelScope outperformed the others, …

Comparative Analysis Of Eye-Metric Algorithms For Code Comprehension In Introductory Cs Course, Noushin Gauhar

College of Graduate Studies: Theses & Dissertations

Eye-tracking technology offers a non-intrusive way to study cognitive processes by tracking where and how long individuals look. In computer science education, it provides valuable insights into how students understand source code—a task that requires intense visual and mental effort. This study investigates how eye-tracking can reveal differences in code comprehension strategies among students in an Introductory Programming course. By analyzing three key metrics—dwell time, gaze entropy, and the K coefficient—the research explores how students engage with code. Dwell time indicates cognitive focus on specific code elements, the K coefficient measures attentional shifts between scanning and focused reading, and gaze …

Enhancing Adhd Diagnosis In College Students Using Multimodal Integration Of Nicats And Iva-2 Tools, Rushmila Shabneen

College of Graduate Studies: Theses & Dissertations

This study explores enhanced methods for accurately identifying Attention Deficit Hyperactivity Disorder (ADHD) indicators in college students. ADHD, a neurodevelopmental disorder, impacts attention, impulse control, and emotional regulation, often leading to academic and social difficulties. Many students remain undiagnosed due to symptom overlap with stress and other factors. Traditional tools like the IVA-2 assess behavioral responses but may not fully capture ADHD complexity. This research integrates IVA-2 data with multimodal metrics from the Non-Intrusive Classroom Attention Tracking System (NiCATS), which monitors facial expressions, eye movements, and computer interactions. Preliminary results show that combining these tools improves ADHD detection, reduces false …

A Deep Learning Approach For Identifying Malicious Activities In The Industrial Internet Of Things, Mohammed Amin Almaiah, Fuad Ali El-Qirem, Rami Shehab, Khaled Sulieman Momani

Mesopotamian Journal of Computer Science

Data-driven decision-making, real-time connectivity, and automation have transformed industrial operations with the Industrial Internet of Things. However, the integration also introduces substantial cybersecurity vulnerabilities, making IIoT networks a prime target for malicious activities. Cyber threats are evolving and becoming more sophisticated, which makes traditional security mechanisms inadequate. An approach using deep learning to detect malicious activities in IIoT environments is examined. It is investigated whether Deep Feed Forward neural networks, autoencoders, and convolutional neural networks are effective at detecting anomalies and mitigating cyber threats. NSL-KDD and UNSW-NB15 benchmark datasets are used to evaluate the proposed model's accuracy, precision, and detection …

Self-Attention Enhanced Dual Bigru For Arabic Fake News Detection, Baqer M. Merzah, Jafar Razmara, Jaber Karimpour

Mesopotamian Journal of Computer Science

The rapid proliferation of social media platforms has greatly amplified the dissemination of fake news, representing significant obstacles to public trust and evidence-based decision-making, particularly for the Arabic-speaking population. Meeting the challenge of Arabic fake news detection is a problem compounded by the complex morphological nature of the language, as well as limited resources. This study presents a hybrid deep learning framework that integrates two Bidirectional Gated Recurrent Units (BiGRUs) along with an attention mechanism for efficiently detecting misinformation in Arabic news. The method leverages FastText word embeddings for disambiguating the intricate semantics of the Arabic language. The model is …

Simulation Of Annpet For Neural Network Training And System Performance Benchmarking, Pete Martone

Graduate Theses, Dissertations, and Problem Reports (ETD)

Monte Carlo (MC) simulations are often used to provide insight into complex physical phenomenon. They can generate realistic results without the necessity to construct costly and complex apparatus. The ultimate value of these simulations depend on how accurately the physics can be modeled and if a sufficient volume can be produced to account for statistical variance. In medical imaging device development, MC simulations of Positron Emission Tomography (PET) scanners have typically been used to model the performance of new systems. The imaging group in the Department of Radiology have developed a unique detector design for a pre- clinical PET scanner …

Using Visual Prompts To Analyze Ejection Fraction In Echocardiograms, Kevin Reisch

Graduate Theses, Dissertations, and Problem Reports (ETD)

Left ventricular ejection fraction (LVEF) is a critical biomarker for heart failure, but manual estimation from echocardiograms is time-consuming. Artificial intelligence can be used to accelerate this process, allowing clinicians to focus on other critical tasks. Current methods typically train models from scratch on echocardiogram datasets; however, this approach is limited by the scarcity of large medical imaging datasets, which are expensive and difficult to acquire. We present a transfer learning approach that leverages pretrained models from massive datasets, enabling continuous improvement as foundation models advance. Our method employs visual prompting to generate trainable masks for echocardiogram videos, transforming the …

Adaptive Fusion Neural Networks For Sparse-Angle X-Ray 3d Reconstruction, Shaoyong Hong, Bo Yang, Yan Chen, Hao Quan, Shan Liu, Minyi Tang, Jiawei Tian

Electrical & Computer Engineering Faculty Publications

3D medical image reconstruction has significantly enhanced diagnostic accuracy, yet the reliance on densely sampled projection data remains a major limitation in clinical practice. Sparse-angle X-ray imaging, though safer and faster, poses challenges for accurate volumetric reconstruction due to limited spatial information. This study proposes a 3D reconstruction neural network based on adaptive weight fusion (AdapFusionNet) to achieve high-quality 3D medical image reconstruction from sparse-angle X-ray images. To address the issue of spatial inconsistency in multi-angle image reconstruction, an innovative adaptive fusion module was designed to score initial reconstruction results during the inference stage and perform weighted fusion, thereby improving …