Bài báo quốc tế
Lĩnh vực nổi bật
Trí tuệ nhân tạo
1.062
Internet vạn vật
23
An toàn bảo mật
11
Blockchain
3
Tìm kiếm theo
Năm xuất bản
Lĩnh vực bài báo
Kho tri thức
/
Bài báo quốc tế
Sắp xếp theo
Liên quan
Integrating CSR and ESG into Sustainable Development Strategies: Practical Solutions for Businesses in the Era of Climate Change
Vũ Thị Hạnh TâmIn the context of escalating climate change, integrating Corporate Social Responsibility (CSR) and Environmental, Social, and Governance (ESG) criteria into sustainable development strategies has become decisive for business survival and growth. Despite strengthened awareness of their importance, many companies, particularly in emerging economies like Vietnam, still struggle with effectively integrating these elements into business operations. This study aims to evaluate the impact of six key factors on CSR and ESG integration effectiveness while proposing practical solutions for businesses in the era of climate change. Using a mixed-methods approach with a sample of 270 enterprises in Vietnam, results show that leadership commitment (β = 0.398) and climate change impacts (β = 0.254) are the two most influential factors, followed by regulatory framework, stakeholder pressure, customer awareness, and financial resources. The regression model explains 64.5% of the variance in CSR and ESG integration effectiveness. The results of this study will serve as a scientific basis for developing a supportive policy framework and sustainable business strategy, affirming that in the context of climate change, integrating CSR and ESG is not only an ethical responsibility but also a smart business strategy to enhance resilience while creating long-term value.
Năm:2026
|
Chủ đề: Trí tuệ nhân tạo
|
Nhà xuất bản: Proceedings of the 8th International Conference on Corporate Social Responsibility and Sustainable Development, Springer Proceedings in Business and Economics
SADL: Sampling, Deliberation, and Pseudo-Labeling for In-Context Learning in Compositional Visual Question Answering
Đặng Hoàng LongLarge vision-language models (LVLMs) offer a novel capability for performing in-context learning (ICL) in Visual QA. When prompted with a few demonstra- tions of image-question-answer triplets, LVLMs have demonstrated the ability to discern underlying patterns and transfer this latent knowledge to answer new questions about unseen images without the need for expensive supervised fine-tuning. However, designing effective vision-language prompts, especially for compositional questions, remains poorly understood. Adapting language-only ICL techniques may not necessarily work because we need to bridge the visual- linguistic semantic gap: Symbolic concepts must be grounded in visual content, which does not share the syntactic linguistic structures. This paper introduces SADL, a new visual-linguistic prompting framework for the task. SADL revolves around three key components: SAmpling, Deliberation, and Pseudo-Labeling of image-question pairs. Given an image-question query, we sample image-question pairs from the training data that are in semantic proximity to the query. To address the compositional nature of questions, the deliberation step decom- poses complex questions into a sequence of subquestions. Finally, the sequence is progressively annotated one subquestion at a time to generate a sequence of pseudo-labels. We investigate the behaviors of SADL under OpenFlamingo on large-scale Visual QA datasets, namely GQA, GQA-OOD, CLEVR, and CRIC. The evaluation demonstrates the critical roles of sampling in the neighborhood of the image, the decomposition of complex questions, and the accurate pairing of the subquestions and labels. These findings do not always align with those found in language-only ICL, suggesting fresh insights in vision-language settings.
Năm:
|
Chủ đề: Khoa học dữ liệu và trí tuệ nhân tạo
|
Nhà xuất bản: Discover Artificial Intelligence
VIVID: A Culturally Grounded Benchmark Exposing the Figurative Language Gap in Vietnamese NLP
Đỗ Trần TúWe present VIVID (Vietnamese Idioms for Validation and Interpretation Depth), the first systematic benchmark for evaluating culturally grounded figurative language understanding in Vietnamese. VIVID comprises 1,636 idioms and proverbs annotated with five complexity traits (literal expressions, pragmatic nuances, Sino-Vietnamese terms, uncommon vocabulary, folk knowledge) and seven semantic themes. We establish an evaluation framework com- bining generative and discriminative tasks, proposing an LLM-as-a-Judge approach with aspect-based prompting validated against human judgment (Cohen’s κ = 0.792). Evaluating eight state-of-the-art models reveals critical gaps: Vietnamese-specialized models drastically underperform multilingual systems (VinaLLaMA-7B: 0.13 vs. GPT-4o: 2.46), and even top models achieve less than 50% correctness on average. Notably, few-shot prompting does not universally improve performance, with GPT-4o exhibiting degradation due to stylistic overfitting. Our analysis exposes systematic failures including literal over-interpretation, lexical gaps, and pragmatic flattening, demonstrating that current models lack cultural competence for nuanced figurative interpretation. VIVID provides an essential tool for advancing figurative language understanding in culturally rich contexts. We release codes and datasets at https://github.com/ReML-AI/VIVID.
Năm:
|
Chủ đề: Trí tuệ nhân tạo
|
Nhà xuất bản:
A characterization of generalized Levitin–Polyak well-posedness for parametric symmetric strong vector quasi-equilibrium problems and its application
Nguyễn Văn HưngThis paper aims to investigate the Levitin–Polyak (LP)-wellposedness
of parametric symmetric strong vector quasi-equilibrium
problems. Firstly, we consider symmetric strong vector quasiequilibrium problems under perturbations. Secondly, we study
the concept of upper semicontinuity in the setting of variable
conic structures for vector-valued mappings and explore their
properties. Thirdly, we establish the LP-well-posedness and generalized LP-well-posedness for these problems under appropriate
conditions. Furthermore, employing the Hausdorff measure to
assess non-compactness, we investigate the metric characterization
of generalized LP-well-posedness for parametric symmetric
strong vector quasi-equilibrium problems. As a final application,
we delve into the LP-well-posedness of symmetric strong vector
quasi-variational inequality problems. The results in this paper are
novel and enhance several key findings in the existing literature. To
illustrate these results, we present multiple examples.
Năm:2026
|
Chủ đề: Toán học và thống kê
|
Nhà xuất bản: OPTIMIZATION
An Interactive Learning and Research Platform for Real-Time Adaptive Feedback Cancellation Systems
Nguyễn Minh TiếnAcoustic feedback is a persistent challenge in audio closed-loop sys-tems such as hearing aids and public address systems, occurring when the sound emitted from the loudspeaker is recaptured by a microphone. The acoustic feed-back phenomenon limits stable gain and causes howling under some circum-stances. A common and effective solution is adaptive feedback cancellation (AFC), which uses adaptive filters to cancel the feedback signal, enhancing sta-bility. A key difficulty for AFC is the inherent correlation between the loud-speaker signal and the desired input (especially when the input signals are spec-trally colored, like speech, music, etc.), which leads to biased estimates of the feedback path. To counter this, methodologies such as the prediction error method (PEM) and its variant decorrelate the input signals of the adaptive filter, improving estimation robustness. The PEM integrated with a range of adaptive algorithms, including NLMS, IPNLMS, H-NLMS, APA, and IPAPA, has been proposed in the literature, each balancing convergence speed, computation, and steady-state error. Since implementing and comparing these algorithms in a unified, real-time context is still challenging, we introduce a comprehensive MATLAB/Simulink platform that provides an end-to-end AFC system. This plat-form enables users to seamlessly switch between various AFC algorithms, toggle pre-filtering, and monitor key performance metrics, such as normalized misalignment (MIS) and added stable gain (ASG) in real-time. This interactive environment serves as both a powerful educational tool and an efficient testbed for vali-dating novel AFC techniques.
Năm:
|
Chủ đề: Trí tuệ nhân tạo
|
Nhà xuất bản:
Toward Robust Malware Detection: A Survey of Datasets, Techniques, and Practical Challenges
Huỳnh Trọng ThưaThe increasing sophistication of malware has diminished the effectiveness of traditional signature-based detection. While Machine Learning (ML), Deep Learning (DL), and Large Language Models (LLMs) have improved malware classification, real-world systems continue to struggle with evasion attacks, temporal drift, and class imbalance. This study reviews the advancements in robust malware detection, focusing on benchmark datasets, detection methods, and operational constraints. Public datasets - EMBER2018, SOREL-20M, MalDICT, MOTIF, and EMBER2024 - are assessed for scale, label quality, and reproducibility. This paper contributes: (i) a Robust Malware Evaluation Protocol (RMEP) for consistent benchmarking under low False-Positive Rates (FPR) (≤ 0.1%) with temporal splits, and (ii) a Dataset-Task-Robustness (DTR) matrix for systematic comparison, offering practical guidance for reproducible malware-detection research. Future efforts should focus on broader multi-platform benchmark coverage, explicit analysis of robustness–accuracy trade-offs, interpretable language-assisted detection pipelines, and privacy-preserving collaborative learning frameworks.
Năm:2026
|
Chủ đề: Khoa học máy tính và thông tin
|
Nhà xuất bản: Engineering, Technology and Applied Science Research (ETASR)
Transfer Learning with Particle Swarm Optimization for Durian LeafDisease Image Classification
Trần Nguyễn Phi HùngTimely durian leaf disease detection is critical for Vietnam’s agricultural productivity, yet traditional methods remainlabor-intensive and error-prone. This study proposes a hybrid pipeline integrating deep transfer learning with binary ParticleSwarm Optimization (PSO) for efficient disease classification. Three lightweight backbones, such as MobileNetV3-Large,EfficientNet-B0, and EfficientNetV2-B0 to extract 128-dimensional features from a real-world Vietnamese durian dataset(2595 images, 6 classes), which PSO then prunes u sing five-fold support vector machine (SVM) cross-validation fitness.The optimized subsets were evaluated across five machine learning (ML) classifiers, achieving up to 92.6% test accuracywith a modest improvement over the baseline while reducing dimensionality. PSO-selected features demonstrated thepotential for accelerated inference and interpretable agricultural diagnostics on resource-constrained devices.
Năm:2026
|
Chủ đề: Khoa học máy tính và thông tin
|
Nhà xuất bản: Applied Fruit Science
A Dual-Loss Vision Transformer Framework for Robust Iris Recognition
Trần Văn HữuNowadays, the iris has become a major instrument, promising to replace fingerprints in identification. In addition, the iris is also a promising biomarker for non-invasive pre-diagnostic problems. To improve the iris recognition ability, in this study, we propose to use Vision Transformer (ViT) combined with Cross-Entropy Loss and Triplet Loss to support the classification. Different from traditional training methods, we conduct a comparison between the ViT model trained from scratch and the ViT-B/16 model fine-tuned from pre-trained weights. The results show that, in the context of limited training data, the fine-tuned model from ViT-B/16 gives significantly higher performance and converges faster than the model trained from scratch. The system is evaluated on various iris databases, including CASIA-Iris-Thousand, CASIA-Iris-Lamp, and CASIA-Interval, demonstrating good stability and generalization ability. The proposed method achieves classification accuracy ranging over 99%, confirming the effectiveness of combining two loss functions and leveraging features from the pretrained ViT model.
Năm:2026
|
Chủ đề: Trí tuệ nhân tạo
|
Nhà xuất bản:
Reasoning-Centric Fake News Detection: A Comprehensive Survey of Architectures, Benchmarks, and Open Challenges
Nguyễn Thanh SơnThe proliferation of misinformation, exacerbated by Generative Artificial Intelligence and Large Language Models, necessitates a paradigm shift from pattern-matching to reasoning-based detection systems. While Transformer-based models and Graph Neural Networks have advanced the field, they often lack interpretability and robustness against adversarial attacks. This survey presents a systematic review of Reasoning-Centric Fake News Detection methodologies for the last five years. Unlike conventional surveys that categorize works by architecture, we propose a novel four-phase evolutionary framework that traces the transition from surface-level semantics to multi-view reasoning paradigms. We introduce a multi-axis taxonomy that unifies content semantics, dynamic propagation topologies, and cross-modal consistency. Furthermore, we critically analyze the “performance saturation” phenomenon (reaching 99% on static benchmarks such as FakeNewsNet), contrasting it with the degradation observed in low-resource languages and in the face of evolving events. Finally, we propose an Integrated Multi-View Reasoning Framework, a five-layer pipeline designed to harmonize automated detection with human-in-the-loop verification, and outline a future roadmap focused on neuro-symbolic reasoning and adversarial resilience in the Artificial Intelligence vs Artificial Intelligence era.
Năm:2026
|
Chủ đề: Khoa học máy tính và thông tin
|
Nhà xuất bản: IEEE Access
UEP-IQ: A Deployable Multi-Modal Phishing Detection Framework
Huỳnh Trọng ThưaPhishing attacks increasingly span multiple surfaces (URLs, spoofed emails, malicious PDFs, and image/QR-based lures), making single-modality detectors brittle in real deployments. We propose UEP-IQ, a unified multi-modal phishing detection framework that combines (i) lightweight Random Forest (RF) classifiers for structured artifacts (URL lexical/structural features, email header anomalies, and PDF metadata descriptors) with (ii) a vision module using EfficientNet-B0 as a frozen backbone and an RF head for phishing screenshots and QR-code misuse. We curate and harmonize heterogeneous public data sources into a consolidated benchmark covering four major phishing vectors, and design modality-specific preprocessing pipelines including a QR to URL linkage to exploit cross-modal signals. Across held-out test sets, UEP-IQ achieves strong per-modality performance (e.g., 93% accuracy for URLs, 99% for email headers and PDFs, and 85–88% for images/QR), outperforming common classical baselines and offering a favorable accuracy–latency trade-off compared to heavier end-to-end models. We further analyze the aggregation layer via fusion sensitivity and provide a learned fusion alternative to justify decision combining beyond heuristics. Finally, we validate deployability through a web application and browser extension, achieving sub-200 ms end-to-end latency on a CPU backend. These results indicate that a production-oriented hybrid architecture can provide scalable, accurate, and practical defense against contemporary multi-surface phishing campaigns.
Năm:2026
|
Chủ đề: Trí tuệ nhân tạo
|
Nhà xuất bản:
Design and Experimental Evaluation of a Hierarchical LoRaMESH-Based Sensor Network with Wi-Fi HaLow Backhaul for Smart Agriculture
Chu Văn CườngLarge-scale smart agriculture requires reliable and energy-efficient wireless connectivity to support distributed environmental sensing across wide rural areas. However, existing low-power wide-area network (LPWAN) technologies often face limitations in scalability, reliability, or infrastructure dependency when deployed in large agricultural fields. This study presents the design and experimental evaluation of a hierarchical sensor network architecture that integrates LoRaMESH for multi-hop sensing communication and Wi-Fi HaLow as a sub-GHz backhaul for data aggregation and cloud connectivity. In the proposed system, LoRaMESH forms intra-cluster sensor networks using a lightweight controlled flooding protocol, while Wi-Fi HaLow provides long-range IP-based connectivity between cluster gateways and a central access point. A real-world deployment covering approximately 2.5km×1km
of agricultural area was implemented to evaluate the performance of the proposed architecture. Experimental results show that the LoRaMESH network achieves packet delivery ratios above 90%
across one to three hops, with average end-to-end delays between 10.6
s and 13.3
s. The Wi-Fi HaLow backhaul demonstrates high reliability within short to medium distances, reaching 99.5%
packet delivery ratio at 50 m and 89.68%
at 200 m. Energy measurements further indicate that the sensor nodes consume only 21.19μA
in sleep mode, enabling long-term battery-powered operation suitable for agricultural monitoring applications. These results indicate that the proposed hierarchical architecture is a feasible connectivity option for the tested large-scale agricultural sensing scenario. Because no side-by-side LoRaWAN or NB-IoT benchmark was conducted on the same testbed, the results should be interpreted as a field validation of the proposed architecture rather than as a direct experimental demonstration of superiority over alternative LPWAN systems.
Năm:2026
|
Chủ đề: Kỹ thuật điện tử và viễn thông
|
Nhà xuất bản: Sensors
Generation Of A New Dataset For An Attack Detection System Towards 5G Network Infrastructure Security
Nguyễn Huy TrungThis paper presents the design and generation of a novel
high-fdelity intrusion detection dataset specifcally targeting 5G core
control-plane attacks. The dataset is constructed using an Open5GSbased testbed integrated with my5G-RANTester, enabling realistic simulation of benign UE registration and advanced authentication-layer
attacks, including MAC failure, SQN desynchronization, replay, brute
force, NAS message manipulation, and denial-of-service scenarios. From
raw packet captures, 25 protocol-aware features are engineered, combining flow-level statistics with entropy-based and sequence-consistency
indicators that reflect 5G-AKA signaling logic. To validate the dataset’s
effectiveness, multiple machine learning models—ranging from Decision
Trees to ensemble methods such as Random Forest and XGBoost—are
evaluated using Accuracy, F1-score, and cross-validation metrics under class imbalance conditions. Experimental results demonstrate that
ensemble models achieve near-perfect classifcation performance with
strong generalization capability, highlighting the discriminative power
of semantic-aware features. The fndings confrm that context-aware feature engineering is essential for reliable intrusion detection in virtualized
5G core infrastructures
Năm:2026
|
Chủ đề: Trí tuệ nhân tạo
|
Nhà xuất bản:
A Two-Stage Agent-based Framework for Network Attack Detection And Categorization in IoT
Nguyễn Huy TrungThis paper presents a method for detecting network attacks targeting resource-constrained IoT devices through the deployment of a lightweight software agent directly on such devices. The proposed agent is cross-platform, capable of being installed on heterogeneous IoT devices, and is designed to collect operational data from the device— including system calls, memory usage, CPU usage, process identifiers (PiD), process hashes, packet captures (PCAP), open ports, bandwidth utilization, and system messages. The agent incorporates a data filtering mechanism to eliminate records that do not exhibit abnormal behavior, thereby reducing processing overhead. The collected data is then analyzed and processed to identify potential network attacks. The proposed solution enables the acquisition of both system-level and network-level data from resource-limited IoT devices, facilitating efficient attack detection and significantly reducing the likelihood of successful cyberattacks on such devices.
Năm:2026
|
Chủ đề: Trí tuệ nhân tạo
|
Nhà xuất bản:
PROPOSED DATA ROAMING NETWORK SIMULATION ENVIRONMENT FOR 5G NETWORK
Nguyễn Huy Trung5G networks are rapidly being deployed, providing high speed and low latency, promoting many new applications. However, the deployment faces several technical challenges. The complex 5G architecture, with virtualisation and network partitioning, increases the attack surface, exposing security vulnerabilities such as insufficient partition isolation and backwards compatibility with older network generations (such as 4G and 3G networks). The roaming environment is a significant security weakness, potentially exposing information and impersonating user subscribers. Since 5G is the foundation for critical future applications, simulating the 5G Roaming environment is necessary to research and test information security solutions. In this paper, we propose a 5G network data roaming simulation model with basic functions for research and testing information security solutions. The model has been tested and evaluated for performance in a laboratory environment to ensure that its parameters are as close as possible to those of an actual 5G network deployment.
Năm:2026
|
Chủ đề: Trí tuệ nhân tạo
|
Nhà xuất bản:
Incremental Real Time Dynamic Multiobjective Path Planning
Lê Minh HóaReal-time path planning is a critical challenge in artificial intelligence, especially in robotics and gaming, directly impacting operational efficiency and user experience. Traditional pathfinding algorithms like A* typically optimize for a single objective and face difficulties in continuously changing dynamic environments that require multiobjective optimization. To address these limitations, the Real-Time Dynamic Multiobjective (RDMO) algorithm was introduced, based on an enhanced A* framework, enabling balanced optimization of multiple objectives and rapid adaptation to environmental changes. However, RDMO still struggles with the overhead of full replanning after each change, particularly in environments with frequent localized updates. This study proposes an improved architecture, I-RDMO, which integrates the incremental search principle of the D* Lite algorithm to reduce replanning costs by updating only the affected regions. Experimental results in simulated environments demonstrate that I-RDMO maintains high processing speed, reduces latency, and improves efficiency compared to traditional methods. This solution is well-suited for applications in real-time strategy games, autonomous robotics, and interactive systems requiring fast response in dynamic settings.
Năm:2026
|
Chủ đề: Trí tuệ nhân tạo
|
Nhà xuất bản:
A Workflow-Oriented Architecture Integrating Large Language Models for Automated Multi-Platform Content Management
Nguyễn Tất ThắngThe rapid growth of social media platforms has significantly increased the complexity of managing digital content across multiple channels. Organizations and content creators increasingly require automated systems that can coordinate publishing, interaction handling, and contextual response generation across heterogeneous platforms. Existing content management tools often provide limited workflow integration and insufficient support for intelligent interaction management. This study proposes a workflow-oriented architecture for automated multi-platform content management that integrates large language models (LLMs) with workflow automation mechanisms. The proposed system is built upon a unified conceptual framework termed CLAW (Content–LLM–Automation–Workflow), which coordinates content resources, contextual information, AI-based response generation, and workflow orchestration within a single operational structure. The
architecture combines centralized backend services, workflow automation, and LLM-based contextual interaction capabilities to support synchronized publishing, multi-account management, and automated response handling across social media platforms. A functional prototype was implemented to demonstrate the feasibility of the proposed architecture. The results show that integrating LLM-based contextual reasoning with workflow orchestration enables more coherent, scalable, and adaptive content management operations. The proposed system provides a practical foundation for future research and development of AIenabled multi-platform content management systems
Năm:2026
|
Chủ đề: Kỹ thuật thông tin
|
Nhà xuất bản: Ingénierie des Systèmes d’Information
Development of an Offline RAG Chatbot for Answering Food Hygiene and Safety Questions Based on Vietnamese Legal Frameworks
Nguyễn Tất ThắngThis paper presents the design, development, and implementation of an offline chatbot system specialized in answering food safety-related questions, relying entirely on Vietnamese legal documents. The system employs Retrieval-Augmented Generation (RAG) to ensure accurate and contextually relevant responses without internet dependency, a critical feature for low-connectivity environments. Key highlights include robust Vietnamese language support, a flexible vector database using Chroma for seamless legal content updates, and the integration of Qwen2.5:7B-Instruct-Q4_0 as the local language model, selected after comparative testing against DeepSeek-R1, Gemma3:1B, and Mistral. Embeddings are generated using BAAI/bge-small-en-v1.5. By processing Vietnamese queries and retrieving from a localized knowledge base, the chatbot delivers reliable guidance to stakeholders such as food producers, traders, and consumers. Evaluations demonstrate high accuracy in Vietnamese Q&A, stable offline operation, and adaptability to evolving regulations, with discussions on limitations and future enhancements.
Năm:2026
|
Chủ đề: Kỹ thuật thông tin
|
Nhà xuất bản: Journal of Research, Innovation and Technologies
Pathways to work engagement in university lecturers: The role of work-life enrichment and life-work enrichment
Trần Thị Bích HằngTeaching staff are central to universities, critically shaping the effectiveness of educational delivery and outcomes. Their roles extend beyond mere instruction. They play a pivotal role in influencing student engagement, motivation, and overall academic success. The enrichment of teaching staff must focus not only on quantity and quality but also on fostering enthusiasm, passion, and professional engagement. This research examines the concurrent impact of work-life enrichment (WLE) and life-work enrichment (LWE) on job satisfaction (JS), life satisfaction (LS), and work engagement (WE) among university lecturers. Drawn on an integration between role theory, social exchange theory, and conservation of resources theory, a conceptual model was proposed and validated by data obtained from a survey of 308 lecturers working at 20 domestic universities in Vietnam, and partial least squares structural equation modeling (PLS- SEM) was used for empirical analysis. Research results indicate that both WLE and LWE of university lecturers positively affect their JS and LS, which in turn enhance their WE. Additionally, JS has a positive influence on LS. While LWE has a positive effect on work engagement, WLE shows no significant effect. The research results offer important contributions to theory as well as suggest implications for higher education institutions to nurture their employee engagement.
Năm:2026
|
Chủ đề: Trí tuệ nhân tạo
|
Nhà xuất bản: Humanities and Social Sciences Letters
Gamifying Human Behavior: How Gamification Drives Consumer Stickiness in E-Commerce
Nguyễn Thị Thanh NhànhObjective: Despite the growing integration of gamification in digital commerce, its impact
on consumer stickiness remains underexplored, particularly in emerging markets. This
study develops and empirically tests a framework examining how specific gamification
elements in e-commerce platforms—badge upgrades, random rewards, and gamified
design—affect consumer stickiness through perceived value (hedonic and utilitarian) and
social interaction. The research aims to clarify the mechanisms through which gamification
enhances customer loyalty and continued platform engagement in the Vietnamese context.
Methods: A questionnaire-based survey was conducted with 310 consumers who had
participated in gamified activities on e-commerce platforms in Vietnam. The study
integrates Partial Least Squares Structural Equation Modeling (PLS-SEM) to examine
linear relationships and Artificial Neural Networks (ANN) to capture nonlinear interactions
within the proposed model. This dual-stage analytical approach enhances the robustness and
predictive power of the findings.
Results: The findings show that gamified design and badge upgrades positively influence
both perceived hedonic and utilitarian values, while random rewards significantly affect
perceived hedonic value only. Social interaction is significantly influenced by gamified
design but not by badge upgrades or random rewards. Perceived value and social interaction,
in turn, contribute to consumer stickiness on e-commerce platforms.
Conclusion: The study confirms that different gamification elements generate distinct
effects on consumer perceptions and stickiness. By highlighting the mediating roles of
hedonic and utilitarian values as well as social interaction, the research contributes to the
literature on smart e-commerce and gamification. The findings suggest that businesses
should strategically design gamification features that simultaneously enhance functional
benefits and experiential enjoyment to strengthen long-term customer retention on digital
platforms.
Năm:2026
|
Chủ đề: Trí tuệ nhân tạo
|
Nhà xuất bản: International Journal of Supply and Operation Management
Ultra-Wide Mid-Infrared Supercontinuum Generation in As2S3 Hexagonal-Core Photonic Crystal Fiber
Dang Van TrongThis work introduces a newly designed As2S3-based photonic crystal fiber structure optimized for generating broadband mid-infrared supercontinuum under low peak power excitation. The hexagonal-core configuration enhances both dispersion and nonlinear characteristics by adjusting structural parameters. Two optimized models, F1 (Λ = 1.0 µm, f = 0.35) and F2 (Λ = 2.0 µm, f = 0.3), are investigated under femtosecond pulse pumping. Fiber F1 supports all-normal dispersion and generates a broadband spectrum from 1.8 to 6.5 µm at 3 kW using a 210 fs pulse. In contrast, fiber F2 yields an anomalous regime SC spectrum spanning 1.5-18 µm under 10 kW peak power and an 85 fs pulse. These findings confirm the applicability of the proposed fibers for MIR applications, such as molecular sensing, spectroscopic analysis, and food quality monitoring.
Năm:2026
|
Chủ đề: Vật lý
|
Nhà xuất bản: Crystal Research and Technology
Digital Transformation in Higher Education as an Organizational Phenomenon: A PRISMA-Based Bibliometric Review
Hoàng Thanh TùngThis study undertakes a systematic review and bibliometric analysis to examine how
digital transformation and innovation in higher education have been conceptualized and
investigated through perspectives related to organizational change and capability development.
Following the PRISMA 2020 protocol, 891 peer-reviewed journal articles published between
2015 and 2025 were retrieved from the Scopus database and analyzed using VOSviewer and
Biblioshiny. The analysis reveals a clear transition from early technology-centered and
adoption-oriented studies toward a more integrated research landscape in which organizational
change, institutional transformation, governance, and adaptive capacity increasingly shape
discussions of digital transformation in higher education. Overall, the findings provide a
structured overview of the field and offer a conceptual foundation for future research seeking
to advance understanding of digitally enabled organizational capabilities in higher education
systems.
Năm:2026
|
Chủ đề: Trí tuệ nhân tạo
|
Nhà xuất bản: Studies in Science of Science
Multi-View Facial Expression Recognition with AU-Guided Heatmaps
Đỗ Hồng QuânFacial Expression Recognition (FER) is fundamental to affective computing and human-computer interaction, enabling systems to infer emotional states from facial cues. Current deep learning approaches rely solely on raw facial images and standard augmentation but lack explicit incorporation of facial anatomical knowledge. This forces models to discover expressiondiscriminative regions from scratch rather than leveraging established facial behavior research. To address this gap, we propose MV-FER, a multi-view framework that integrates spatial priors from the Facial Action Coding System (FACS). MV-FER employs three data representations: original facial images, horizontally flipped images with random erasing for view consistency, and AU-guided heatmaps generated by mapping landmarks to established Action Unit areas providing explicit spatial guidance toward expression-relevant facial regions. The three views are unified within a training paradigm that encourages learning of discriminative features while maintaining robustness against pose variations and partial occlusions. Benchmark experiments demonstrate that MV-FER achieves enhanced recognition accuracy over state-of-the-art approaches with similar architectures. These results confirm the effectiveness of our multi-view strategy, with progressive accuracy gains as augmented and AU-guided heatmap views are incorporated.
Năm:2026
|
Chủ đề: Trí tuệ nhân tạo
|
Nhà xuất bản:
HiSeR: Hybrid Item Encoding for Transformer-based Sequential Recommendation
Nguyễn Thị YếnSequential recommendation systems play a crucial role in predicting the next item a user is likely to interact with. Although recent models have achieved remarkable progress in modeling interaction sequences, their ability to exploit and integrate semantic information from user reviews remains limited. Moreover, effectively incorporating these heterogeneous features into a unified representation for sequential modeling remains an open challenge. In this study, we propose HiSeR: Hybrid Item Encoding for Transformer-based Sequential Recommendation that introduces a novel approach to multimodal information fusion. Unlike prior methods, HiSeR employs a Hybrid Item Encoder capable of deeply synthesizing features from both item IDs and textual reviews, resulting in rich and unified semantic representations for each item. These hybrid representations are then fed into a Transformer-based sequential encoder to model the temporal dynamics of user preferences. Extensive experiments on Amazon Product Reviews dataset demonstrate that HiSeR significantly outperforms several state-of-the-art baselines, confirming the effectiveness of the proposed multimodal fusion strategy in enhancing sequential recommendation quality.
Năm:2026
|
Chủ đề: Trí tuệ nhân tạo
|
Nhà xuất bản:
Intelligent UAV Positioning and Fair Power Allocation via DRL in Cloud-Impaired HAP-to-UAV FSO/RF Systems
Nguyễn Quốc HuyThis paper proposes an adaptive resource allocation strategy for a non-terrestrial network (NTN) architecture, where a high-altitude platform (HAP) provides free-space optical (FSO) backhaul links, and unmanned aerial vehicles (UAVs) serve as radio-frequency (RF) access nodes for end users. The goal is to ensure a fair and efficient allocation of transmission capacity among users. However, atmospheric turbulence and cloud dynamics in FSO links can severely degrade system performance. To address this issue, this work integrates deep reinforcement learning (DRL) to optimize UAV positioning in real time, enabling UAVs to avoid cloud-blocked regions and maintain stable connectivity for users. Based on channel states and user rate requirements, the proposed method decouples subchannel assignment from power allocation and incorporates DRL-based UAV trajectory control, forming a comprehensive optimization framework. The numerical results demonstrate that the proposed approach significantly improves system throughput, maintains user fairness, and reduces transmit power requirements compared to conventional methods.
Năm:
|
Chủ đề: Trí tuệ nhân tạo
|
Nhà xuất bản: