Digital Commons Network^™

Llms Are Bad Judges. So Use A Classifier Instead., Jack Kieffaber, Kimo Gandall, Steven Foster, Kenny Mclaren

SMU Law Review Forum

Large Language Models suffer from prompt variance—meaning they’ll give you totally different legal answers depending on how you phrase your question. Jonathan Choi demonstrated this recently when he asked ChatGPT five legal questions, each rephrased 2,000 times, and watched as the bot spat out different answers every time. When you tell somebody that AI is going to replace the judge, the lawyer, and the legal system in the next twenty years, Choi’s article has become the go-to rebuttal; it’s the crown jewel of the “AI bad” genre.

Choi’s absolutely right that LLM’s are bad judges. And, if every AI was …

Intelligent Decision-Making Systems In Smart Greenhouses, Muso Berdiyor Ugli Allanov

Chemical Technology, Control and Management

Smart greenhouses offer a solution to sustainable food production under climate uncertainty, yet their management often depends on fixed rules or human intuition. This study proposes an intelligent decision-making framework that integrates optimization, simulation, and a neural set into a self-learning system. By generating “conditionally real data” through simulation and evolutionary algorithms, the system can predict microclimatic changes and optimize control of water, energy, and nutrients. Continuous digital feedback enables adaptive, data-efficient operation even with limited real data. Experimental results demonstrate reduced resource use and improved yield stability, advancing the development of autonomous and resilient greenhouse ecosystems.

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 …

Being The “Other”: How Teaching Internationally Impacts Teacher Candidates’ Culturally Sustaining Practices, Jeanne Carey Ingle, Emily Spitzman, Sarah Thomas

Bridgewater Review

In an increasingly globalized world, preparing teachers to work with culturally and linguistically diverse (CLD) learners has become a critical component of teacher education programs. The recently revised Massachusetts Department of Elementary and Secondary Education (DESE) Candidate Assessment of Performance (CAP) guidelines emphasize the need for teacher candidates to demonstrate growth in their culturally sustaining pedagogy (DESE, 2025). However, traditional approaches to cultural responsiveness in teacher preparation often fall short of comprehensive, practice-oriented methods that foster connections to classroom practice (Gulya & Fehérváry, 2023).

Bridgewater Review Volume 43 Number 2 2025

Bridgewater Review

No abstract provided.

Explainable Sentiment Analysis With Deepseek-R1: Performance, Efficiency, And Few-Shot Learning, Donghao Huang, Zhaoxia Wang

Research Collection School Of Computing and Information Systems

Large language models (LLMs) have transformed sentiment analysis, yet balancing accuracy, efficiency, and explainability remains a critical challenge. This study presents the first comprehensive evaluation of DeepSeek-R1—an open-source reasoning model—against OpenAI’s GPT-4o and GPT-4o-mini. We test the full 671B model and its distilled variants, systematically documenting few-shot learning curves. Our experiments show DeepSeek-R1 achieves a 91.39% F1 score on 5-class sentiment and 99.31% accuracy on binary tasks with just 5 shots, an eightfold improvement in few-shot efficiency over GPT-4o. Architecture-specific distillation effects emerge, where a 32B Qwen2.5-based model outperforms the 70B Llama-based variant by 6.69 percentage points. While its reasoning …

Efficient Integration Of External Knowledge To Llm-Based World Models Via Retrieval-Augmented Generation And Reinforcement Learning, Chang Yang, Xinrun Wang, Qinggang Zhang, Qi Jiang, Xiao Huang

Research Collection School Of Computing and Information Systems

World models achieve remarkable success in predicting future states and planning in complex environments and Large Language Models (LLMs) serve as promising foundation to build general world models. However, their performances are usually constrained by the limited external knowledge to specific environments. Existing research attempts to enhance LLM-based world models through prompting or fine-tuning approaches, which are either requiring human knowledge or computationally extensive. Therefore, we introduce Retrieval-Augmented World Models (RAWM), a novel framework that leverages retrieval-augmented generation to efficiently integrate the external knowledge to LLM-based world models. Our main contributions are threefold: (i) We introduce a memory system and …

One Planner To Guide Them All! Learning Adaptive Conversational Planners For Goal-Oriented Dialogues, Huy Dao, Lizi Liao

Research Collection School Of Computing and Information Systems

Goal-oriented dialogues, such as recommendation and negotiation, often require balancing multiple, conflicting objectives. Existing methods typically involve training separate models for specific combinations of objectives, leading to computational and scalability issues. In this work, we aim to develop a new dialogue policy method that can adapt to varying objective preferences at inference time without retraining. This raises several challenges in terms of both (1) optimization strategy and (2) knowledge utilization. To address these, we propose a novel learning framework, Preference Adaptive Dialogue Policy Planner (PADPP), for multi-objective goal-oriented dialogues. Specifically, to tackle the former, we introduce a novel policy optimization …

Gerunds In Irish Sign Language: An Exploratory Analysis, Zaid Mohammed

Doctoral

In sign languages (SLs), e.g., American Sign Language (ASL), gerunds help express the meaning of ongoing actions and activities conveyed by verbs (Klima & Bellugi, 1979, p. 295). In spoken languages, a gerund is a form of the verb that functions as a noun and expresses the state of being or an ongoing action (Maekelberghe, 2020). Gerunds are complex constructions as they have nominal-verbal nature (Travis, 2005, p. 321) and morphological similarity with present participle or progressive aspect (Siegel, 1998). This research is motivated by the challenging nature of distinguishing gerund constructions at the morphological level, and research in sign …

In Companies We Trust: Cheap Talk Or Genuine Virtue?, Bishal Bc, Thuy Simpson

American Business Review

This paper examines how corporate narratives support either the impression management hypothesis or the incremental information hypothesis by analyzing the frequency of trust words used in the Management Discussion and Analysis (MD&A) section of 10-K filings. By directly linking trust language to actual fines and violations enforced by more than 400 federal, state, and local regulatory agencies in the U.S. from 2010 through 2023, we find that firms using more trust words in their MD&A tend to incur higher monetary penalties. We also document a significant link between increased use of trust words and poorer firm performance. Additional analyses reveal …

Identifying Sex- And Gender-Specific Endocrinological, Lifestyle, Psychosocial, And Socio-Cultural Targets For Alzheimer's Disease Prevention In Africans: The Female Brain Health And Endocrine Research In Africa (Fember-Africa) Project, Chinedu Momoh, Benard Aliwa, Lukoye Atwoli, Karen Blackmon, Edna Bosire, Samuel Gitau, Harrison Kaleli, Ciru Kamanda, Linda Khakali, Rachel Maina, Peter Mativo, Sylvia Mbugua, Zul Merali, Kendi Muchungi, Anne Njogu, Douglas Nyankira, Alice Ondieki, Catherine Onyancha, Stanely Onyango, Jasmit Shah, Sheena Shah, Cynthia Smith, Dilraj Sokhi, Sheila Waa, Thomas Thesen

Brain and Mind Institute

Dementia rates are rising globally, with the burden increasing most rapidly in low- tomiddle-income countries. Despite this, research into Alzheimer’s disease and relateddementias (ADRD) among African populations remains limited, with existing modelsbased on Western cohorts that overlook sex-, gender-, and ancestry-specific factors.The Female Brain Health and Endocrine Research in Africa (FemBER-Africa) project,hosted at the Brain and Mind Institute, Aga Khan University, Kenya, will establish adeeply phenotyped cohort of 250 African individuals across the ADRD spectrum. Itwill assess sex-specific risk factors linked to ethnicity, lifestyle, and endocrinologicalvariables using fluid-based biomarkers (blood and saliva), neuroimaging (magneticresonance imaging and positron emission tomography), and …

Afterword, Kristina A. Hesbol

The Rural Educator

No abstract provided.

Hey, Bruh . . . I Mean, Your Honor—The Evolution (Or Devolution) Of Legal Communications In Modern Times, Michael J. Fleck

Northern Illinois University Law Review

The ability to effectively communicate, whether such communication is oral or written, is a primary tool in any lawyer’s toolbox. How we communicate is a reflection on our level of professionalism, both as individual attorneys, and as the legal body at large. It is how we convey ideas with our client, with our colleague, and with the court. It is how we persuade through legal reasoning. There is no doubt that the style of communication has evolved over time. One need only to read legal opinions from a century ago to see what was acceptable as proper communication. The standard …

A Systematic Review And Integrative Framework Of Psychotherapy Microprocess: Linking The Science Of Psychological Interventions With The Art Of Moment-To-Moment Practice, Xiaochen Luo, Alytia A. Levendosky

Counseling Psychology

Despite advancements in psychotherapy research on effectiveness and critical therapy processes, there remains a significant gap between the science and the art of psychotherapy, specifically on how to understand what to do moment-to-moment with each patient. A burgeoning research literature addresses this question by examining psychotherapy microprocesses, which typically referred to within-session changes of therapy processes, aiming to bridge psychotherapy research with clinical practice. In this pre-registered systematic review, we reviewed 86 empirical quantitative studies examining observational psychotherapy microprocesses over 35 years. We extracted 28 microprocess constructs across six categories (affective/emotional, behavioral, cognitive, relational/interpersonal, linguistic, and movement), four key methodological …

Communicating The Green Transition - Practices And Competencies For Engineers, Paul Talbot, Una Beagon, Stephen Harper, Janet Mckennedy

Research Outputs: 2025-Present

The green transition is driving demand for engineering expertise in sectors such as renewable energy, construction, transport, and industrial decarbonisation. Projects are complex and interdisciplinary, and communication plays a central role. This report presents findings from 20 focus groups with 103 engineers across Europe examining how engineers communicate in green transition project contexts. Participants described a wide range of communicative tasks involving coordination, documentation, stakeholder engagement, and public communication.Engineers work across spoken, written, and digital modes, shifting between formal and informal styles, and adapting to varied audiences. They must navigate fragmented tools and workflows, unclear goals and roles, jargon pitfalls, …

Understanding Behavioral Engagement In Project-Based Learning: Key Influencing Factors, Maria Coney G. Pallones

English Faculty Publications

Project-Based Learning (PBL) is widely acknowledged for fostering student engagement; however, limited research explores how its specific design features influence behavioral engagement, defined as students’ active participation, persistence, and effort, particularly in early education settings. This study investigates what key factors, as perceived by teachers, influence students' behavioral engagement during PBL, and how these factors support motivation through the lens of Self-Determination Theory (SDT). Semi-structured interviews were purposively conducted with elementary teachers experienced in implementing PBL, and the data were analyzed using thematic analysis. Three key design features emerged as critical to sustaining behavioral engagement: voice, choice, and ownership, which …

Evaluating The Role Of Investor Sentiment Dynamics In Tadawul Stock Market Using Hybrid Wang-Mendel Fuzzy Logic: A Bridge To Neutrosophic Uncertainty, Abdullah Alawajee, Mohd Tahir Ismail, S. Al Wadi, Jamil Jaber

Neutrosophic Sets and Systems

No abstract provided.

Distance Spectrum And Energy In Single Valued Neutrosophic Graphs, A.S. Sasipriya, Lakshmi Madireddy

Neutrosophic Sets and Systems

No abstract provided.

A Model For Failure Mode And Effects Analysis Based Single Valued Neutrosophic Sets, Hossein Sayyadi Tooranloo, Reihaneh Hafizi Atabak

Neutrosophic Sets and Systems

No abstract provided.

Linguistic Neutrosophic Sets With Application To Group Decision-Making To Enhance The Work Effectiveness Evaluation Of University Counselors, Li Wang

Neutrosophic Sets and Systems

The evaluation of university counselors' work effectiveness in the new era aims to assess their performance in areas such as ideological and political education (IPE), psychological counseling, academic support, and daily management. The evaluation typically includes counselors' performance in guiding students’ thoughts, promoting mental health, crisis intervention, and career planning. A scientific and comprehensive evaluation system can improve the quality of counselors' work and promote students' holistic development. The work effectiveness evaluation of university counselors is multiple-attribute group decision-making (MAGDM). Recently, Exponential TODIM (ExpTODIM) and TOPSIS approaches have been introduced to address MAGDM. The 2-tuple linguistic neutrosophic sets (2TLNSs) have …

Design And Implementation Of Data Quality Controls In The Eq-Daphnie Study: Insights From The Pilot Phase And 15-Country Analysis, Fatima Al Sayah, Hilary Short, Juan M. Ramos-Goñi, Rosalie Viney, Erica I. Lubetkin, Mathieu F. Janssen, Jeffrey A. Johnson

Publications and Research

Objective

The EQ-DAPHNIE (EuroQol Data for Assessment of Population Health Needs and Instrument Evaluation) project is a large, multi-country survey initiative designed to generate population norms and enable comparative research using self-reported health measures. This paper describes the quality control processes and summarizes data quality metrics from the United Kingdom (UK) pilot and full implementation across 15 countries.

Methods

Representative samples were recruited via Dynata, an online survey panel provider, using quota sampling by age, sex, income, community setting, and language (where applicable). The UK pilot (n = 3012) informed survey refinements ahead of full rollout (n = 68,411). Quality …

Bridging Cultures, Building Communities: The Role Of Academic Libraries In Fostering Multicultural Engagement And Digital Inclusion, Emily Chan, Catarina Lei

Proceedings of the IATUL Conferences

This paper explores the evolving role of academic libraries in small, multicultural settings. Using the University of Saint Joseph (USJ) in Macao as a case study, we examine how the library acts as a catalyst for interdisciplinary collaboration. By bridging the gap between diverse faculty departments—such as History, Sound Engineering, and Design—the library facilitates innovative projects that preserve local heritage and promote Sustainable Development Goals (SDGs). This document outlines two specific case studies: the "First Global Typhoon Macau 1874" exhibition and the "OttoSonic" sound installation.

Sparking Collaborative Connections Through Immersive Technologies, Jacky Wong

Proceedings of the IATUL Conferences

NIE Library created an innovative space called LIBytes to connect people and ideas through immersive technologies— Virtual Reality (VR), Augmented Reality (AR), Mixed Reality(MR), and AI robots—and to widen equitable access to digital experiences. LIBytes is both a space and a service: we co-design with departments to integrate immersive learning into teaching, research, and outreach.

For instance, we partner across NIE to build shared experiences:

• English Department – Digital Literature: VR for immersive narrative exploration and perspective-taking, with multimodal discussion.
• Asian Languages & Cultures Department – Digital Storytelling: co-creating VR stories that surface linguistic and cultural perspectives and support authentic …

Adaptive Vs. Traditional Learning: Long-Term Knowledge Retention -A Literature Review, Amal Altalhi, Chaitali Bonke, Khaled Alblowi

Information Systems Faculty Publications

Adaptive learning, a personalized educational approach, has appeared as a substitute paradigm to conventional teaching methodologies. Opposed to instruction-based learning, adaptive learning prepares learning content in a way that corresponds to individual learner needs, increasing engagement and knowledge retention. The present study has been conducted to review the literature to evaluate the influence of adaptive learning systems on long-term knowledge retention as compared to their traditional counterparts. Real-time feedback, spaced repetition, and scaffolded content can reduce cognitive load and enhance the learning experience, as they are considered highly effective tools. Several studies have shown that retention improves through the use …

Singapore Awakened: How Success – And Flourishing – Shape Family, Seow Hon Tan

Research Collection Yong Pung How School Of Law

Assoc. Prof. (Dr.) Tan Seow Hon delivered the keynote address at Cultivate SG’s second annual conference, “Unfiltered – The Family on Trial”, on 17 November 2025. In her speech titled “Singapore Awakened: How Success – and the Alternative of Flourishing – Shape Family”, Dr Tan reflects on the “Singapore Dream”, the narrow mindset of success in contrast with the concept of flourishing, and how these impact marriage and family. She concludes by offering some thoughts on how to move from success to flourishing.

Comorbidities And Emotions - Unpacking The Sentiments Of Pediatric Patients With Multiple Long-Term Conditions Through Social Media Feedback: A Large Language Model-Driven Study, Temidayo Oluwalade, Hossein Ahmadi, Lin Huo, Richard Sharpe, Shang-Ming Zhou

School of Nursing and Midwifery

ObjectivesThe emotional and psychological challenges faced by children with multiple long-term conditions (MLTCs) remain underexplored. This study aimed to analyze sentiments and emotions expressed by this vulnerable population and their caregivers on social media, assess the effects of comorbidities and the COVID-19 pandemic on emotional well-being.MethodsNarratives from the Care Opinion platform (2008–2023) were analyzed by a model called CoEmoBERT, developed using the large language model, distilroberta-base transformer model. The CoEmoBERT-based sentiment analysis categorized emotions into “Positive”, “Negative”, and “Neutral,” with further refinements into specific emotions such as “Sad,” “Fear”, “Satisfied” etc. through pretraining and transferring process. Comorbidity associations with emotions …

Hypersoft Set With Neutrosophic Sets For Quality Evaluation Of Visual Communication Design Professional Courses, Jing Yao, Xingli Wang

Neutrosophic Sets and Systems

The quality evaluation of the Visual Communication program is a comprehensive assessment of course design, teaching effectiveness, and student learning outcomes. Through evaluation, the rationality of course content, achievement of teaching objectives, and effectiveness of teaching methods can be understood. It helps instructors improve their teaching strategies, ensuring that students acquire the necessary professional skills and theoretical knowledge. Additionally, evaluation results reflect whether the course keeps pace with industry trends and meets the needs of society and the market. Student feedback provides important references for course optimization, driving continuous improvement in teaching quality, and ultimately cultivating outstanding design talents who …

A Novel Method For Examining Autistic Children’S Comprehension Of Individual Words Produced Within Delayed Echolalia: A Proof-Of-Concept Pilot Study, Janine Mathee-Scott, Grace Corrigan, Emily Lorang, Zachary Hesse, Jennifer Johnson, Courtney E. Venker

Speech Pathology and Audiology Faculty Research and Publications

Delayed echolalia, or the repetition of previously heard speech, is often observed in the expressive language of autistic children. Relatively little is known about how the production of delayed echolalia fits within the overall picture of autistic children’s language ability, including receptive language. To date, no empirical studies have tested autistic children’s comprehension of individual words within their delayed echoes. The present study aimed to establish proof-of-concept for a novel method of examining children’s comprehension of individual words that they produce in their own delayed echoes. Using natural language sampling combined with parent report, we identified instances of delayed echolalia …

Language, Law, And Wealth Destruction In Puerto Rico, Scott M. Brown, Daniel J. Hall

Vanderbilt Journal of Transnational Law

No abstract provided.

When It Comes To Scientific Information Extraction And Llms, Less Is More, Sameer Shaik

Theses and Dissertations from DePaul University

The scientific literature continues to expand rapidly, making manual extraction of structured scientific facts increasingly impractical. Traditional Machine Learning and Natural Language Processing (NLP) pipelines require large expert-annotated datasets, which are costly to produce. Novel Large Language Models (LLMs) face challenges in long-context scientific reasoning, hallucinations, and entity linking. This thesis investigates ELSIE-Blob, a domain-aware preprocessing method that segments scientific articles into compact text “blobs” containing components of entity relations (here, polymer names, melting point indicators, and numerical values). We test whether blob-based input allows lightweight, consumer-hardware-accessible LLMs to extract polymer–melting point (polymer–Tm) pairs accurately without training data. Experiments using …