Digital Commons Network^™

Rendering Data Unlearnable By Exploiting Llm Alignment Mechanisms, Ruihan Zhang, Jun Sun

Research Collection School Of Computing and Information Systems

Large language models (LLMs) are increasingly trained on massive, heterogeneous text corpora, raising serious concerns about the unauthorised use of proprietary or personal data during model training. In this work, we address the problem of data protection against unwanted model learning in a realistic blackbox setting. We propose Disclaimer Injection, a novel data-level defence that renders text unlearnable to LLMs. Rather than relying on model-side controls or explicit data removal, our approach exploits the models’ own alignment mechanisms: injecting carefully designed alignment-triggers to prevent effective learning. Through layer-wise analysis, we find that finetuning on such protected data induces persistent activation …

Train In Vain: Functionality-Preserving Poisoning To Prevent Unauthorized Use Of Code Datasets, Yuan Xiao, Yuchen Chen, Jiaming Wang, Wei Song, Jun Sun, Shiqing Ma, Yanzhou Mu, Juan Zhai, Chunrong Fang, Jin Song Dong, Zhenyu Chen

Research Collection School Of Computing and Information Systems

The widespread availability of large-scale code datasets has accelerated the development of code large language models (CodeLLMs), raising concerns about unauthorized dataset usage. Dataset poisoning offers a proactive defense by reducing the utility of such unauthorized training. However, existing poisoning methods often require full-dataset poisoning and introduce transformations that break code compilability. In this paper, we introduce FunPoison, a functionality-preserving poisoning approach that injects short, compilable weak-use fragments into executed code paths. FunPoison leverages reusable statement-level templates with automatic repair and conservative safety checking to ensure side-effect freedom, while a type-aware synthesis module preserves type correctness, suppresses static-analysis warnings, and …

Itimo: An Llm-Empowered Synthesis Dataset For Travel Itinerary Modification, Zhuoxuan Huang, Yunshan Ma, Hongyu Zhang, Hua Ma, Zhu Sun

Research Collection School Of Computing and Information Systems

Addressing itinerary modification is crucial for enhancing the travel experience as it is a frequent requirement during traveling. However, existing research mainly focuses on fixed itinerary planning, leaving modification underexplored due to the scarcity of shape need-to-modify itinerary data. To bridge this gap, we formally define the itinerary modification task and propose a general pipeline to construct the corresponding dataset, namely iTIMO. This pipeline frames the generation of shape need-to-modify itinerary data as an intent-driven perturbation task. It instructs large language models to perturb real-world itineraries using three operations: REPLACE, ADD, and DELETE. Each perturbation is grounded in three intents: …

Dual-Diffusional Generative Fashion Recommendation, Mingzhe Yu, Lei Wu, Qianru Sun, Yunshan Ma

Research Collection School Of Computing and Information Systems

Personalized generative recommender systems have emerged as a promising solution for fashion recommendation. However, existing methods primarily rely on implicit visual embeddings from historical interactions, which often contain preference-irrelevant information and result in insufficient user behavior modeling. Moreover, these models typically generate only item images, providing limited interpretability. To address these limitations, we propose DualFashion, a Dual-Diffusional Generative Fashion Recommendation Architecture that jointly models image and text modalities for personalized and explainable recommendation. DualFashion adopts a dual-diffusion Transformer with image and text branches, where structured attribute-level captions and visual outfit information are jointly used as conditioning signals to model user …

Avadclip: Audio-Visual Collaboration For Robust Video Anomaly Detection, Peng Wu, Wanshun Su, Guansong Pang, Yujia Sun, Qingsen Yan, Peng Wang, Yanning Zhang

Research Collection School Of Computing and Information Systems

With the increasing adoption of video anomaly detection in intelligent surveillance domains, conventional visual-only detection approaches often struggle with information insufficiency and high false-positive rates in complex environments. To address these limitations, we present a novel weakly supervised framework that leverages audio-visual collaboration for robust video anomaly detection. Capitalizing on the exceptional cross-modal representation learning capabilities of Contrastive Language-Image Pretraining (CLIP) across visual, audio, and textual domains, our framework introduces two major innovations: an efficient audio-visual fusion that enables adaptive cross-modal integration through lightweight parametric adaptation while maintaining the frozen CLIP backbone, and a novel audio-visual prompt that dynamically enhances …

Robust Graph Learning On The Web: Challenges, Methods, And Applications, Ao Xiang, Yang Liu, Guansong Pang, Yuanhao Ding, Hezhe Qiao, Dawei Cheng, Qing He

Research Collection School Of Computing and Information Systems

Graph learning is transforming web intelligence, powering applications from recommender systems to anomaly detection. However, most existing approaches implicitly assume ideal conditions where training and testing data are accurate, complete, and free from manipulation. In reality, web environments rarely exhibit such stability. Dynamic user behavior, incomplete or outdated content, adversarial interference, and sudden distribution shifts can all erode the reliability of even state-of-the-art models, leading to biased or unsafe outcomes. This tutorial provides a comprehensive survey of emerging strategies for robust graph learning on the web. We first present a structured taxonomy of the principal robustness threats specific to web …

Verbalizing Lightgcn: Direct Learning Of Textual Representations From User-Item Interaction Graph Via Llms, Manh-Khanh Ngo Huu, Hady Wirawan Lauw

Research Collection School Of Computing and Information Systems

In this work, we propose VerbaLightGCN, a novel LLM-based recommendation framework that integrates the semantic understanding of LLMs with user-item interaction modeling. Traditional collaborative filtering (CF) models typically embed user and item IDs into a latent space to capture interaction signals. However, pretrained LLMs cannot natively interpret these learned embeddings. To bridge this gap, VerbaLightGCN adopts a CF-as-text paradigm, in which collaborative signals are encoded in textual form and directly learned from the user–item interaction graph, and are then combined with semantic information to construct user and item profiles that function as latent embeddings. Inspired by LightGCN, our method retains …

Divide-And-Conquer: Generating Ui Code From Screenshots, Yuxuan Wan, Chaozheng Wang, Yi Dong, Wenxuan Wang, Shuqing Li, Yintong Huo, Michael R. Lyu

Research Collection School Of Computing and Information Systems

Websites are critical in today’s digital world, with over 1.11 billion currently active and approximately 252,000 new sites launched daily. Converting website layout design into functional UI code is a time-consuming yet indispensable step of website development. Manual methods of converting visual designs into functional code present significant challenges, especially for non-experts. To explore automatic design-to-code solutions, we first conduct a motivating study on GPT-4o and identify three types of issues in generating UI code: element omission, element distortion, and element misarrangement. We further reveal that a focus on smaller visual segments can help multimodal large language models (MLLMs) mitigate …

The Power And Peril Of Awe In Leadership: Transforming Follower Identity And Behavior, Jack Mcguire, Daniel Mcallister, Jochen Menges, David De Cremer

Research Collection Lee Kong Chian School Of Business

Awe is a profound emotion that has captured significant attention within psychological research. While the potential for leaders to inspire awe in followers has received some recognition, systematic research on the nature and effects of awe in leadership—and within organizational contexts more broadly—remains limited. In this article, we offer a conceptual framework that explains the multifaceted and transformative nature of leadership through the power of awe. Specifically, we identify four leader behaviors—charismatic leadership tactics, exceptional performance, problem reframing, and self-sacrificial behavior—that elicit awe among followers. We further propose three variants of awe-inspiring leaders, describing how variation in a leader’s self-construal …

Connected, Traceable And Trusted Identifier In China, Jia Liu, Xiaolei Xia, Lijuan Wang, Shu Wang, Yuanchun Zhou

FORCE 2026

Persistent Identifiers (PIDs) has been recognized as a critical infrastructure for the Open Science. The Chinese National Standard GB/T 32843-2016, Science and Technology Resource Identification was launched in 2016 which specifies the numbering schema of the Science and Technology Resource Identification(CSTR). The standard systematically defines the identification objects, encoding rules, and applications for science and technology resources, establishing a robust technical foundation and operational guidelines for their complete lifecycle management. CSTR Identification Platform（https://www.cstr.cn/en） based on the Standard, providing unique identification service for global science data, papers, patents and other science and technology resources.
Connected
CSTR platform services cover …

[Birds-Of-A-Feather] A Tri-Level Roadmap To Ai-Ready Universities: Strategy, Operations, Deployment, Sarah Forzetting, Rochelle Lundy, Ahmad Pratama, Karryl Kim Sagun Trajano, Antara Chakraborthy, Yasmine Wong

FORCE 2026

This panel discussion by three institutions will offer three approaches to conceptualizing and supporting AI use at the national, local, and individual levels.

Nanyang Technological University will frame the discussion by examining AI as a social good, highlighting how even advanced initiatives such as Singapore’s face equity gaps that risk leaving vulnerable populations behind. The panelist will further outline how information professionals and other key stakeholders can play a critical role in bridging these gaps in the pursuit of widespread public AI literacy. Stanford University Libraries will outline difficulties institutions have in operationalizing guidelines to help researchers bring a project …

Developing An Assessment Framework To Support Critical Evaluation Of Ai-Powered Academic Search Engines, Huajin Wang

FORCE 2026

As AI-powered tools are emerging in all aspects of research, it’s challenging for researchers, students, and librarians alike to assess the benefit, usefulness, and risks associated with a new AI-powered tool or AI-enhanced features of an existing one. To address this challenge, Carnegie Mellon University Libraries is developing a Tool Evaluation Framework to help users understand how AI-powered tools work, initially focused on AI-enhanced academic search engines. By evaluating criteria such as mechanisms for document retrieval, processes for content generation, and the quality of outputs, this framework guides users to carefully consider the strengths and weaknesses of a tool before …

Supporting Fair Practices In Scholarly Publishing With The Editorial Reference Handbook, Susanna-Assunta Sansone, Allyson Lister, Rebecca Taylor-Grant, Matthew Cannon

FORCE 2026

Co-produced by academics and publishers (incl. CUP, Cell Press, EMBO Press, Taylor & Francis, GigaScience Press, OUP, PLOS, Springer Nature), the Editorial Reference Handbook (https://publishers.fairassist.org) assists scholarly publishers in supporting the sharing of digital research objects and in operationalising FAIR research practices by addressing gaps in editorial workflows, policy implementation and stakeholder alignment. The Handbook comprises three interrelated components—a checklist, detailed guidance documentation, and a flowchart—intended primarily for in-house editorial staff while also providing value to reviewers, authors, and service providers.

Beside this practical collaboratively developed product, the Handbook is also a socio-technical pilot to improve the culture …

Ten Years Of Tuwhera: A Commitment To Sustaining Open Access Publishing, Donna Coventry

FORCE 2026

Tucked away, far from the conference circuit, is a part of the world which doesn’t tend to make the headlines when it comes to open access initiatives. New Zealand and by extension, the South Pacific, is a region struggling with a history of colonisation and trying to make systematic change around decolonisation - open access to research about ‘our place in the world’ is vitally important. 

Ten years ago Auckland University of Technology’s library (Te Mātāpuna) made the decision to start hosting Diamond open access journals. Starting with two established journals, the service was named Tuwhera which can be translated …

Asian Open Research Data And Its Potential For Ai: Bridging The Digital Divide Through Strategic Data Sharing, Mark Hahnel, Simon Porter

FORCE 2026

The growing momentum for open research data in Asia intersects with global advances in artificial intelligence (AI). Open data policies and infrastructures are increasingly recognized as critical enablers of research equity, reproducibility, and innovation. This talk examines the current state of open academic data, highlights economic and scientific arguments for its adoption, and explores the transformative potential of Asian open data ecosystems in powering AI-driven discovery. Drawing on global examples such as the Protein Data Bank and emerging health datasets, the discussion positions Asian institutions to leverage open strategies that simultaneously meet compliance mandates, enhance visibility, and accelerate breakthroughs in …

Managing Ai Bot Access To Open Scholarly Infrastructures, Petr Knoth

FORCE 2026

The rapid rise of generative AI has created unprecedented demand for large, high-quality research corpora. Open access repositories and other open scholarly infrastructures have therefore become primary sources for AI bots, because they host research that is not universally reliable, but remains far more evidence-based than most web content. This is both an opportunity and a strain: repositories are now more valuable than ever, but machine traffic brings sustainability, capacity and policy challenges. Repositories may be able to scale, but who should fund that scaling, and under what conditions?

The core dilemma is how to curb abusive high-load bot activity …

Beyond The Launch: Seven Years Of Sustaining An Open Educational Resource After The Grant Ends, Susan Gardner Archambault

FORCE 2026

This presentation shares seven years of lessons from maintaining Project CORA (Community of Online Research Assignments), an award-winning, open-access platform where librarians and educators share adaptable information literacy assignments. Launched in 2016 with grant support and strong community engagement, CORA saw early success: international page views across 150+ countries, 200+ shared assignments and teaching resources, and community-building features like "adapt this assignment" and leaderboards.

But when funding ended in 2018, Project CORA entered what many open projects experience but few discuss publicly: the precarious post-grant phase, when initial funding and enthusiasm fade but infrastructure still needs support. Annual contributions dropped …

[Birds-Of-A-Feather] Understanding And Handling Software Plagiarism In The Age Of Generative Ai, Daniel S. Katz, Scott C. Edmunds

FORCE 2026

Computing and software have supported research since their inception, and continue to play a significant role in knowledge production. However, the means of communicating research methods and results were developed long before computing existed, and the research community lacks best practices for documenting computational research elements transparently, reproducibly, and reusably.
Publishers are now more accepting of the inclusion of software (typically, source code) associated with submitted manuscripts, and many want to support processes to vouch for the integrity of software just as they do for other content, such as ensuring that ethical and legal concerns such as authorship, plagiarism, copyrights …

Diamond Open Access Journals In India: Status, Sustainability And Challenges, Mallikarjun Dora, Kanagasabai K, Raj Kishor Kampa

FORCE 2026

Open-access publishing has transformed scholarly communication in recent decades. The share of OA articles between 2014 and 2024 increased by 26%, from a modest 14% in 2014 to 40% in 2024. OA journals have evolved over time and exist in various models based on funding mechanisms, including the Gold, Green, Hybrid, and Diamond models. While OA publishing emerged in response to scientific inequality, middle- and low-income countries face the barrier of high Article Processing Charges (APC), which hinder them from publishing in those journals (Alperin, 2022). The Article Processing Charges are part of the Gold and Hybrid OA models, where …

Can Ai Help In Systematic Reviews? A Comparative Study Of Manual, Ai-Assisted, And Ai-Dominant Workflows For Librarians And Researchers, Aster Zhao

FORCE 2026

Systematic reviews are essential to evidence-based research but are often time-consuming and labor-intensive, for researchers to conduct, and for librarians to participate in, or teach. The rapid advances in Generative AI and AI-powered research tools, have created a growing interest in whether these technologies can assist—or even automate—parts of the systematic review process. However, questions remain about their effectiveness, reproducibility, and potential bias. This study explores the role of GenAI tools in the early stages of a systematic review: from literature discovery and screening, up to the identification of included studies.

We propose a comparative case study of one systematic …

Genai In Qualitative Data Analysis: Framework-Guided Prompt Engineering In Library Research Practice, Debby R. Wegener

FORCE 2026

As Generative AI (GenAI) tools become increasingly integrated across the research lifecycle, researchers need practical, reproducible methods for the responsible use of these technologies. This presentation will demonstrate a systematic approach to using GenAI for qualitative data analysis through a case study of thematic coding in a library website usability study at the Singapore Institute of Technology.

Drawing on prompt engineering frameworks like CLEAR, this session will illustrate how structured prompts can maintain academic rigour and enhance the reliability of GenAI-assisted analysis. The presentation will walk through the complete workflow, that is, from initial data preparation and tool selection to …

Using Ai-Assisted Programming To Develop Research Services Tools For Research Impact, Open Access Publishing & More, Gary Lee

FORCE 2026

Academic libraries play a vital role in scholarly communication, As research practices become more data‑driven and interdisciplinary, librarians can help scholars by creating and sharing flexible, customizable tools that align with local workflows and user needs.

AI Assisted programming (sometimes called “vibe coding”) offers a new way for librarians without extensive programming knowledge to develop such tools . This allows previously non-expert librarians to go from conceptual goals to working applications by rapid prototyping, experimentation, and roll-out, resulting in service innovation and improvement.

This presentation illustrates how librarians at HKUST have explored vibe coding with tools like GROK, POE, and …

Recognizing And Rewarding Peer Review: Rethinking Research Assessment For Openness, Fairness, And Global Equity, Tung Tung Chan, Bernd Pulverer, Johan Rooryck

FORCE 2026

The Coalition for Advancing Research Assessment (CoARA) Working Group on Recognizing and Rewarding Peer Review has developed a comprehensive framework for reforming how scholarly review is valued within research careers. Our recommendations address a fundamental question: how can peer review, a critical yet often invisible scholarly contribution, be made visible, credited, and meaningfully integrated into research assessment?

Developed through a collaborative effort across 16 European organisations, the Working Group’s outputs offer targeted recommendations for four key stakeholder groups: research performing organizations, research funding bodies, publishers and editors, and individual researchers. These recommendations are structured across five key dimensions:
(1) Openness: …

Open Research Information - How To Support Publishers To Make Metadata Openly Availble, Bianca Kramer

FORCE 2026

Research information, or scholarly metadata, is important for decision making around strategic priorities, distribution of resources, and evaluation of researchers and institutions. It is also used to assess the effect of policies, and to find and assess research results. Open research information (free to access and free to (re)use) is increasingly valued for fair assessment and equitable decision making, and is also important in digital sovereignty.


The Barcelona Declaration on Open Research Information calls on organizations performing, funding and evaluating research to make openness of research information the default, work with services and systems that support and enable open research …

Mapping Themes, Aligning Strategies: Institutional Research Analytics With Bertopic And Genai, Danping Dong, Pin Pin Yeo

FORCE 2026

Libraries and research offices are increasingly tasked with making sense of unstructured institutional data, such as research publications and strategic documents. These insights help inform decision-making and leadership planning. This session shares how SMU Libraries leveraged BERTopic, an open-source topic modeling tool, together with generative AI tools such as ChatGPT and Claude, to extract meaningful insights from faculty publications and support institutional sensemaking.

In one use case, we applied BERTopic to university-authored publications to surface thematic groupings related to research about Asia. Generative AI tools were then used to summarize each cluster into human-readable narratives, enabling research landscape mapping, identification …

Public Acceptance Of Meat-Reduction Policies Across Cultures: Comparing Singapore And Switzerland, Bianca Wassmann, Shu Tian Ng, Mark Chong, Angela K. Y. Leung, Michael Siegrist

Research Collection Lee Kong Chian School Of Business

Purpose – As meat-reduction policies are discussed across the globe, many are met with public resistance. Cultural values may help explain this pushback, yet their role in shaping support for food policy remains poorly understood. This study is the first to apply cultural cognition theory to food policy by examining how cultural worldviews shape the acceptance of meat-reduction interventions in Singapore and Switzerland—two economically developed countries with contrasting cultural profiles. Design/methodology/approach – In an online survey, participants (Singapore: n = 357; Switzerland: n = 495) rated their acceptance of 11 meat-reduction interventions (e.g. taxes, subsidies, labelling). We then analysed to …

Hydpn: A Hybrid Deep Reinforcement Learning, Programming, And Neighborhood Operations Framework For Integrated Scheduling On Parallel Batch Processing Machines, Yuqi Wang, He Luo, Guoqiang Wang, Zhaoxia Wang

Research Collection School Of Computing and Information Systems

Batch processing machines (BPMs) are widely used in industries such as semiconductors, metal processing, and healthcare, where jobs are processed in batches. As production, inventory, and distribution become increasingly integrated to improve efficiency, research on their joint scheduling in parallel BPM environments remains scarce. This paper addresses the integrated scheduling problem in parallel BPMs, involving production, inventory, and distribution stages, with the objective of minimizing total costs. A unified cost-based model is first formulated, applicable to both in-facility and external distribution scenarios. A hybrid algorithm framework, HyDPN, combining deep reinforcement learning, dynamic programming, and neighborhood operations is proposed. Extensive experiments …

On-The-Fly Generation-Quality Enhancement Of Deep Code Models Via Model Collaboration, Weifeng Sun, Naiqi Huang, Meng Yan, Zhongxin Liu, Hongyan Li, Yan Lei, David Lo

Research Collection School Of Computing and Information Systems

The growing prominence of deep code models in automating software engineering tasks is undeniable. However, their deployment encounters significant challenges in on-the-fly performance enhancement, which refers to dynamically improving the performance of deep code models during real-time execution. Conventional techniques, such as retraining or fine-tuning, are effective in controlled pre-deployment scenarios but fall short when adapting to on-the-fly adjustments post-deployment. CodeDenoise, a notable on-the-fly performance enhancement technology, leverages uncertainty-based methods to identify misclassified inputs and applies an input modification strategy to rectify classification errors. While effective for classification tasks, this approach is inapplicable to generative tasks due to two key …

Interfold: Learning Interpretable Diffusion Manifolds Beyond Binary Samples, Alexander Vincent Lewi, Rainer Tan, Shengfeng He

Research Collection School Of Computing and Information Systems

We propose InterFold, a framework for learning and applying interpretable semantic manifolds in latent diffusion models, without requiring binary or paired supervision. Existing methods for semantic editing either rely on limited paired data or uncover only coarse, unsupervised directions that fail to capture user-specific, fine-grained attributes. InterFold addresses these limitations by learning a target attribute manifold in the H-space of diffusion models using only a set of positive, unlabeled examples. To edit a new image, InterFold projects its H-space representation toward this learned manifold through test-time optimization, enabling precise, identity-preserving modifications of complex, non-binary concepts. To make these edits effective …

Videocreator: An Agentic System For Multi-Turn Video Production, Zhengyang Liang, Yan Shu, Cathal Gurrin, Nicu Sebe, Lizi Liao

Research Collection School Of Computing and Information Systems

Recent advances in video generation models enable visually compelling single clips. However, real-world video creation is inherently continuous and iterative: creators refine content over multiple rounds while maintaining narrative, style, and entity consistency. Existing standalone generators are largely stateless and lack memory of previously generated segments, making it difficult to produce a coherent and consistent video project. To address this gap, we present VideoCreator, a unified video agent that integrates generation and understanding with a project-level memory system. VideoCreator leverages understanding capabilities to perform fine-grained analysis of newly produced content and uses persistent memory to retain and reuse prior context …