Rendering Data Unlearnable By Exploiting Llm Alignment Mechanisms,
2026
Singapore Management University
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,
2026
Singapore Management University
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,
2026
Singapore Management University
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,
2026
Singapore Management University
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,
2026
Singapore Management University
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,
2026
Singapore Management University
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,
2026
Singapore Management University
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 …
Unmasking Twitter Bots: An Applied Machine Learning Approach,
2026
Department of Mathematics and Computer Science, Faculty of Science, Beirut Arab University, Lebanon
Unmasking Twitter Bots: An Applied Machine Learning Approach, Rayane El Raba’A, Layal Abu Daher
BAU Journal - Science and Technology
The rapid growth of social networks has led to increased challenges, such as fraud, cyberbullying, and the spread of automated accounts (bots). Detecting anomalies within these networks is essential to maintaining security and trust. This study explored machine learning algorithms: Random Forest, XGBoost, Support Vector Machine (SVM), and Logistic Regression for anomaly detection in social networks, specifically focusing on Twitter bot identification, By applying AI-driven data mining techniques to a dataset of 37,438 Twitter bot accounts dataset, the research evaluates the effectiveness of these models in detecting unusual patterns. XGBoost achieved the highest accuracy (84.9%), with an ROA_AUC of 0.87, …
Enhanced E-Commerce Recommendation Experience With Collaborative Sentiment Analysis And Ranked Content-Based Filtering,
2026
Department of Mathematics and Computer Science, Faculty of Science, Beirut Arab University, Lebanon
Enhanced E-Commerce Recommendation Experience With Collaborative Sentiment Analysis And Ranked Content-Based Filtering, Sandi El Zein, Farah Hamze, Amani Merhi, Lynn Noureddine, Lama Affara
BAU Journal - Science and Technology
The rise of online shopping has made informed purchasing decisions increasingly difficult, as consumers face an overwhelming number of product choices and struggle to manually evaluate specifications and user reviews. This paper presents an AI-powered, sentiment-aware product recommendation tool that effectively aligns user preferences with real-world customer feedback from online reviews. The proposed system utilized Instruct ABSA deep learning models for feature extraction and DeBERTa-v3 for sentiment analysis to turn reviews into interpretable scores that would nominate optimal products. An interactive Rasa-based chatbot interface, TopPickAI, was developed to give a seamless user experience, educate users on product features, and conversationally …
On The Use Of Lorawan For Smart Fishing Applications In The Blue Economy Sector,
2026
Department of Electronics and Telecommunications Engineering, University of Dar es Salaam, P.O. Box 33335, Dar es Salaam., Tanzania
On The Use Of Lorawan For Smart Fishing Applications In The Blue Economy Sector, Eva Shayo
Tanzania Journal of Engineering and Technology (TJET)
The growth of digital technology is expected to transform small-scale fishery sectors, where a need for robust, low-cost, long-range communication networks becomes critical. There exist several technologies that are used in the fishery sector but they are never affordable to small scale fisheries. This study evaluates the feasibility of using low cost Long Range Wide Area Network (LoRaWAN) technology specifically tailored for smart fishing environments to small scale fishery sector. Using simulation, we assess the performance of the key performance metrics including probability of success and energy efficiency under varying device densities and time. During evaluation, we considered end devices …
Multi-Modal Survival Prediction On Breast Cancer Mammograms And Enabling Multi-Institutional Collaborations With Federated Learning,
2026
University of South Florida
Multi-Modal Survival Prediction On Breast Cancer Mammograms And Enabling Multi-Institutional Collaborations With Federated Learning, Nikolas Koutsoubis
USF Tampa Graduate Theses and Dissertations
Cancer remains one of the leading causes of mortality worldwide, and machine learning is well positioned to leverage the large volumes of routinely collected clinical and imaging data to improve patient outcomes. Realizing this potential at scale requires three capabilities that current practice does not yet deliver in combination: models that integrate the multiple data modalities clinicians use, training procedures that respect institutional data-sharing constraints, and tooling that enables sharing of de-identified data when federated approaches are not sufficient. This dissertation develops methods and infrastructure addressing each of these capabilities in the context of breast cancer and oncology more broadly. …
A Scoping Review Of Sycophancy In Large Language Models: Operational And Theoretical Recognition,
2026
Mississippi State University
A Scoping Review Of Sycophancy In Large Language Models: Operational And Theoretical Recognition, Kallen Zhou, Manning Littlejohn, Isabella Garrard
Endeavors: Mississippi State Undergraduate Research Journal
As large language models (LLMs) usage grows across different domains, sycophancy, the tendency for output to align with users, is increasingly being recognized as a primary issue arising from applying LLMs into critical areas. Current research has provided a variety of theoretical definitions, mitigation techniques, and quantification for sycophancy. However, there is little to no consistency across different papers. This scoping review seeks to connect different works on LLM sycophancy by identifying themes in theoretical definitions, measurement methods, and inducement techniques of sycophancy. By analyzing 26 papers (preprints, conference proceedings, and journal articles) from arXiv, ACL Anthology, and Scopus, this …
Adaptive Path Planning For Robotic Arms Integrating Rrt* And Apf,
2026
School of Intelligence Science and Technology, Xinjiang University, Urumqi 830017, China
Adaptive Path Planning For Robotic Arms Integrating Rrt* And Apf, Zhirun Chen, Jie Yuan, Erkenbieke Jia, Ningning Zhang, Chao Liu, Yushan Ye
Journal of System Simulation
To address the issues of large search space, low efficiency, and slow convergence of the RRT* algorithm in 3D path planning of robotic manipulators, an adaptive path planning algorithm integrating RRT* and APF is proposed. In the sampling phase, a Sobol sequence-based obstacle avoidance strategy and an APF adaptive-threshold, goal-biased sampling method are used to improve the quality of sampling points. During the expansion phase, sampling, attractive, and repulsive vectors are integrated, and adaptive weights are designed based on environmental information to generate a resultant force direction, thus enhancing the expansion guidance. For step size control, the obstacle repulsive potential …
Design Of Teaching Model For Simulation Engineering Comprehensive Practical Course,
2026
College of System Engineering, National University of Defense Technology, Changsha 410073, China
Design Of Teaching Model For Simulation Engineering Comprehensive Practical Course, Meigen Huang, Xin Zhou, Tian Jing, Tao Wang
Journal of System Simulation
To cultivate postgraduates' comprehensive engineering problem-solving ability to complex problems, based on Wittrock's generative learning theory and the agile systems engineering paradigm, this paper proposed an overall idea of curriculum construction reform promoted by the integration of teaching and research, independently developed an unmanned swarm attack and defense simulation experimental teaching platform, constructed a comprehensive practical case of "warship defense against unmanned aerial vehicle swarm attacks", and refined a relatively advanced and distinctive teaching model for the simulation engineering comprehensive practical course. This model effectively strengthens postgraduates' engineering problem-solving thinking for complex problems, stimulates their independent innovation ability, elevates …
Dynamic Analysis Of Flexible Mechanism With Lubrication Clearances Considering Uncertain Parameters,
2026
College of Mechanical and Electronic Engineering, Shandong University of Science and Technology, Qingdao 266590, China
Dynamic Analysis Of Flexible Mechanism With Lubrication Clearances Considering Uncertain Parameters, Xiulong Chen, Yi Sun, Aiguo Wang
Journal of System Simulation
This paper conducted a dynamic analysis on a seven-link mechanism suitable for a hybriddriven press, simultaneously considering lubrication clearances, interval parameters, and link flexibility. Based on the Lagrange equations of the first kind, the dynamic equations of the flexible mechanism with lubrication clearances under deterministic parameters were established; interval variables were introduced to establish a dynamic model considering interval parameters, and the Chebyshev interval algorithm and the Runge - Kutta method were used to solve the response interval curves of interval parameters such as clearances value, dynamic viscosity, and cross-sectional area. The results show that compared with the results under …
Energy Management Strategy For Hybrid Electric Buses Considering Vehicle Mass Variation,
2026
School of Traffic & Transportation Engineering, Central South University, Changsha 410075, China
Energy Management Strategy For Hybrid Electric Buses Considering Vehicle Mass Variation, Jinjun Tang, Shuaijie Zhang
Journal of System Simulation
The vehicle mass variation during the operation of buses affects power demand of the vehicle, which can result in poor performance of energy management strategies. To this end, a hybrid electric bus energy management strategy based on proximal policy optimization-adaptive simulated annealing (PPOASA) is proposed. ASA is introduced into PPO to perturb policy parameters according to policy entropy before the policy update, and the perturbed policies are adaptively accepted or rejected by employing the Metropolis criterion, thus improving the exploration capability of the policy and convergence stability. Experimental results show that the proposed method outperforms the charge depleting-charge sustaining (CD-CS) …
Research On Optimization Modeling Method For Eye Tracking In Solfeggio Cognitive Simulation,
2026
School of Electrical and Automation, Nantong University, Nantong 226019, China
Research On Optimization Modeling Method For Eye Tracking In Solfeggio Cognitive Simulation, Kun Zhang, Jiajie Qian, Shuhong Ma, Zengxu Zhao, Yuchen Pan, Yaoqi Tang
Journal of System Simulation
To address the fixation offset problem caused by head movement in music solfeggio teaching simulation and the lack of system-level simulation validation in existing methods, this paper proposed a fixation accuracy optimization method integrating image semantic understanding, temporal trajectory modeling, and solfeggio cognitive simulation. With Vision Transformer as the core, after preprocessing via Mahalanobis distance, sliding window, and region of interest, position offset perception, offset residual regression, and dual-pathway fusion were introduced to achieve offset modeling and correction under unlabeled conditions. Simulation results indicate that the error of this method decreases by 43.9% compared with the original value error; removing …
Exploration And Practice Of Talent Cultivation System For System Modeling And Simulation,
2026
School of Automation Science and Electrical Engineering, Beihang University, Beijing 100191, China
Exploration And Practice Of Talent Cultivation System For System Modeling And Simulation, Shaoping Wang, Chao Zhang, Ni Li, Yong Cui, Yongjia Zhao, Quan Quan
Journal of System Simulation
Focusing on the cultivation of innovative talents in modeling and simulation of aerospace and safety-critical control systems and facing the problems and challenges in the education and teaching of talent cultivation for system modeling and simulation, this paper inherited the "red ambition, soaring far" spirit of simulation researchers in Beihang University, constructed an integrated bachelor, master, and doctoral course system empowered by the latest key technologies, designed open and collaborative aerospace teaching cases, independently developed a control system simulation teaching platform based on Chinese simulation software MWORKS, and constructed a hardware-in-the-loop simulation experimental verification system for C919 aircraft, to …
Exploration Of Talent Cultivation System And Practical Model For Simulation And Optimization Of Intelligent Manufacturing System,
2026
School of Mechanical Science & Engineering, Huazhong University of Science and Technology, Wuhan 430074, China; The National Research Institute for Teaching Materials (Research on Higher Education Teaching Materials for Intelligent Manufacturing and Robotics), Huazhong University of Science and Technology, Wuhan 430074, China
Exploration Of Talent Cultivation System And Practical Model For Simulation And Optimization Of Intelligent Manufacturing System, Xinyu Li, Zheng Duan, Liang Gao, Chunjiang Zhang, Peigen Li
Journal of System Simulation
To address problems such as the insufficient integration of science and education in the talent cultivation system for traditional manufacturing system simulation and optimization, the insufficient integration of industry and education in cultivation goals and approaches, and the lack of full-chain industrial-level practical cultivation means, a "1223" reform scheme for innovative talent cultivation in the intelligent manufacturing system was formed. Research and practice were carried out focusing on the talent cultivation system, cultivation approaches, and practical cultivation resources for the simulation and optimization of the intelligent manufacturing system. Significant outcomes were achieved in aspects of innovative talent cultivation, faculty and …
Exploration Of Online-Offline Integrated Practical Teaching Path Driven By Digital-Intelligent Simulation,
2026
School of Mechanical Engineering, Beijing Institute of Technology, Beijing 100081, China
Exploration Of Online-Offline Integrated Practical Teaching Path Driven By Digital-Intelligent Simulation, Zhen Zuo, Dong Zhang, Zhi Wang, Zhongxin Li
Journal of System Simulation
To address the disconnection between theory and practice, industry and education, and scientific research and teaching in traditional practical teaching, this paper utilized digital-intelligent simulation to empower practical teaching, promoted the collaborative linkage between national-level science and education platforms and high-quality industrial research and development platforms, and constructed a vehicle-oriented "CAE simulation, autonomous driving simulation, and performance testing simulation" practical system. Guided by the constructivist learning theory, this paper innovated the practical model by integrating online and offline approaches, built a competency-oriented practical education ecosystem through competition-education integration, and formed a new "competency-led, industry-oriented, and project-driven" paradigm of …
