Search Results (603)

Under rapid urbanization and ecological transformation, balancing authenticity preservation with adaptive reuse presents a major challenge for industrial heritage landscapes. This study investigates the Dagushan Iron Mine in Anshan, China’s first large-scale open-pit iron mine and once the deepest in Asia, which is currently undergoing ecological backfilling that threatens its core landscape morphology and spatial integrity. Using a mixed-method approach combining archival research, spatial documentation, qualitative interviews, and expert evaluation through the Analytic Hierarchy Process (AHP), we construct a cross-validated evidence chain to examine how evidence-based industrial landscape heritage values can inform low-intervention digital narrative strategies for off-site learning. This study contributes theoretically by reframing authenticity and integrity under ecological transition as the traceability and interpretability of landscape evidence, rather than material survival alone. Evaluation involving key stakeholders reveals a value hierarchy in which historical value ranks highest, followed by social and cultural values, while scientific–technological and ecological–environmental values occupy the mid-tier. Guided by these weights, we develop a four-layer value-to-narrative translation framework and an animation design pathway that supports curriculum-aligned learning for off-site students. This study establishes an operational link between evidence chain construction, value weighting, and digital storytelling translation, offering a transferable workflow for industrial heritage landscapes undergoing ecological restoration, including sites with World Heritage potential or status. Full article

With the acceleration of digital transformation, open-source software has become a fundamental component of modern smart grids and other critical infrastructures. However, the complex dependency structures of open-source ecosystems and the continuous emergence of vulnerabilities pose substantial challenges to software supply chain security. In power information networks and cyber–physical control systems, vulnerabilities in open-source components integrated into Supervisory Control and Data Acquisition (SCADA), Energy Management System (EMS), and Distribution Management System (DMS) platforms and distributed energy controllers may propagate along the supply chain, threatening system security and operational stability. In such application scenarios, large language models (LLMs) often suffer from limited semantic accuracy when handling domain-specific security terminology, as well as deployment inefficiencies that hinder their practical adoption in critical infrastructure environments. To address these issues, this paper proposes KD-SecBERT, a domain-specific semantic bidirectional encoder optimized through multi-level knowledge distillation for open-source software supply chain security in smart grid applications. The proposed framework constructs a hierarchical multi-teacher ensemble that integrates general language understanding, cybersecurity-domain knowledge, and code semantic analysis, together with a lightweight student architecture based on depthwise separable convolutions and multi-head self-attention. In addition, a dynamic, multi-dimensional distillation strategy is introduced to jointly perform layer-wise representation alignment, ensemble knowledge fusion, and task-oriented optimization under a progressive curriculum learning scheme. Extensive experiments conducted on a multi-source dataset comprising National Vulnerability Database (NVD) and Common Vulnerabilities and Exposures (CVE) entries, security-related GitHub code, and Open Web Application Security Project (OWASP) test cases show that KD-SecBERT achieves an accuracy of 91.3%, a recall of 90.6%, and an F1-score of 89.2% on vulnerability classification tasks, indicating strong robustness in recognizing both common and low-frequency security semantics. These results demonstrate that KD-SecBERT provides an effective and practical solution for semantic analysis and software supply chain risk assessment in smart grids and other critical-infrastructure environments. Full article

Fault-tolerant control in safety-critical industrial systems demands adaptive responses to equipment degradation, parameter drift, and sensor failures while maintaining strict operational constraints. Traditional model-based controllers struggle under these conditions, requiring extensive retuning and dense instrumentation. Recent safe multi-agent reinforcement learning (MARL) frameworks with control barrier functions (CBFs) achieve real-time constraint satisfaction in robotics and power systems, yet assume comprehensive state observability—incompatible with sensor-hostile industrial environments where instrumentation degradation and contamination risks dominate design constraints. This work presents a safety-aware multi-agent deep reinforcement learning framework for adaptive fault-tolerant control in sensor-lean industrial environments, achieving formal safety through learned implicit barriers under partial observability. The framework integrates four synergistic mechanisms: (1) multi-layer safety architecture combining constrained action projection, prioritized experience replay, conservative training margins, and curriculum-embedded verification achieving zero constraint violations; (2) multi-agent coordination via decentralized execution with learned complementary policies. Additional components include (3) curriculum-driven sim-to-real transfer through progressive four-stage learning achieving 85–92% performance retention without fine-tuning; (4) offline extended Kalman filter validation enabling 70% instrumentation reduction (91–96% reconstruction accuracy) for regulatory auditing without real-time estimation dependencies. Validated through sustained deployment in commercial beverage manufacturing clean-in-place (CIP) systems—a representative safety-critical testbed with hard flow constraints (≥1.5 L/s), harsh chemical environments, and zero-tolerance contamination requirements—the framework demonstrates superior control precision (coefficient of variation: 2.9–5.3% versus 10% industrial standard) across three hydraulic configurations spanning complexity range 2.1–8.2/10. Comprehensive validation comprising 37+ controlled stress-test campaigns and hundreds of production cycles (accumulated over 6 months) confirms zero safety violations, high reproducibility (CV variation < 0.3% across replicates), predictable complexity–performance scaling ($R^2=0.89$), and zero-retuning cross-topology transferability. The system has operated autonomously in active production for over 6 months, establishing reproducible methodology for safe MARL deployment in partially-observable, sensor-hostile manufacturing environments where analytical CBF approaches are structurally infeasible. Full article

Universities face a dual challenge: supporting student mental health while equipping staff to respond effectively. To address this, we co-designed and embedded the “You Matter, Prioritize Your Wellbeing” intervention within the university curriculum using a participatory action research framework. The program was developed through co-design workshops and a student needs survey, piloted across six undergraduate courses, and refined into a scalable Facilitator’s Toolkit. Data were collected from co-design workshop participants (n = 23 staff, n = 7 students), student survey respondents (n = 109), academic facilitators’ interview (n = 5), and student post-pilot feedback (n = 61). Purposive sampling was used for co-design workshops, and convenience sampling for both surveys. A mixed-methods approach was employed: qualitative data were analysed using reflexive thematic analysis, and quantitative data using descriptive statistics. Evaluation showed strong student engagement, with 82% planning proactive self-care. Academic facilitators reported enhanced confidence and competence in facilitating wellbeing conversations, valuing the structured approach for normalizing the topic while maintaining professional boundaries. Synchronous delivery and authentic facilitator sharing were perceived as especially impactful. Despite systemic barriers, all facilitators expressed commitment to continued use. This study presents a practical, scalable model for a whole-of-university approach to wellbeing, moving beyond siloed support services to foster a proactive culture of care in higher education. Full article

Background/Objectives: Retinal optical coherence tomography (OCT) is essential for diagnosing ocular diseases, yet developing high-performing multiclass classifiers remains challenging due to limited labeled data and the computational cost of self-supervised pretraining. This study aims to address these limitations by introducing a curriculum-based self-supervised framework to improve representation learning and reduce computational burden for OCT classification. Methods: Two ensemble strategies were developed using progressive masked autoencoder (MAE) pretraining. We refer to this curriculum-based MAE framework as CurriMAE (curriculum-based masked autoencoder). CurriMAE-Soup merges multiple curriculum-aware pretrained checkpoints using weight averaging, producing a single model for fine-tuning and inference. CurriMAE-Greedy selects top-performing fine-tuned models from different pretraining stages and ensembles their predictions. Both approaches rely on one curriculum-guided MAE pretraining run, avoiding repeated training with fixed masking ratios. Experiments were conducted on two publicly available retinal OCT datasets, the Kermany dataset for self-supervised pretraining and the OCTDL dataset for downstream evaluation. The OCTDL dataset comprises seven clinically relevant retinal classes, including normal retina, age-related macular degeneration (AMD), diabetic macular edema (DME), epiretinal membrane (ERM), retinal vein occlusion (RVO), retinal artery occlusion (RAO), and vitreomacular interface disease (VID) and the proposed methods were compared against standard MAE variants and supervised baselines including ResNet-34 and ViT-S. Results: Both CurriMAE methods outperformed standard MAE models and supervised baselines. CurriMAE-Greedy achieved the highest performance with an area under the receiver operating characteristic curve (AUC) of 0.995 and accuracy of 93.32%, while CurriMAE-Soup provided competitive accuracy with substantially lower inference complexity. Compared with MAE models trained at fixed masking ratios, the proposed methods improved accuracy while requiring fewer pretraining runs and reduced model storage for inference. Conclusions: The proposed curriculum-based self-supervised ensemble framework offers an effective and resource-efficient solution for multiclass retinal OCT classification. By integrating progressive masking with snapshot-based model fusion, CurriMAE methods provide high performance with reduced computational cost, supporting their potential for real-world ophthalmic imaging applications where labeled data and computational resources are limited. Full article

Assessment practices in Romanian primary education remain largely recall-based, despite curriculum expectations that prioritize reasoning, metacognition, and inclusive learning processes. This conceptual–analytical study examines the structural misalignments between curriculum goals, classroom assessment cultures, and national evaluation systems, highlighting their impact on learning equity and cognitive development. Drawing on international frameworks (OECD, UNESCO), national assessment data, and Romanian pedagogical literature, the analysis identifies three systemic gaps: curriculum–assessment misalignment, assessment–instruction misalignment, and a mismatch between equity-oriented policies and classroom practice. To address these challenges, the article proposes the ECEI Framework, an integrated developmental model that combines principles of cognitive education, metacognitive strategy development, inclusive pedagogy, and formative assessment. The framework introduces four categories of indicators—cognitive, metacognitive, inclusive, and assessment—designed to support teachers in observing and evaluating learning processes more effectively in diverse classrooms. Discipline-based illustrations in mathematics, reading, and science demonstrate how innovative assessment practices can make students’ thinking visible through authentic tasks, learning progressions, and multimodal response pathways. The findings suggest that developmental and inclusive assessment is essential for improving learning outcomes and reducing socio-economic disparities in primary education. Implementing the ECEI Framework requires targeted teacher training, coherent curriculum–assessment alignment, and system-level support to ensure sustainable changes in instructional practice. Full article

Personalized face image generation is essential for Artificial Intelligence-Generated Content (AIGC) applications such as personalized digital avatars and user-customized media creation. However, existing diffusion-based approaches still suffer from insufficient identity consistency and limited text editability. In this work, we propose CoFaDiff, a diffusion-based face generation framework designed for coordinating identity consistency and text-driven editability. To enhance identity consistency, our method integrates a dual-encoder scheme that jointly leverages CLIP and ArcFace to capture both semantic and discriminative facial features, incorporates a progressive curriculum learning strategy to obtain more robust identity representations, and adopts a hybrid IdentityNet–IPAdapter architecture that explicitly models facial location, pose, and corresponding identity embeddings in a unified manner. To enhance text-driven editability, we introduce three complementary optimization strategies: First, long-prompt fine-tuning is employed to reduce the model’s dependency on identity conditions. Second, a semantic alignment loss is incorporated to regularize the influence of identity embeddings within the semantic space of the pretrained diffusion model. Third, during classifier-free guided sampling, we modulate the strength of the identity condition by stacking different numbers of zero-valued identity tokens, enabling users to flexibly balance identity consistency and text editability according to their needs. Experiments on FFHQ and IMDB-WIKI demonstrate that CoFaDiff achieves superior identity consistency and text editability compared to recent baselines. Full article

Building Information Modeling (BIM) has become a core competency in architectural practice, prompting increasing efforts to integrate BIM into design education. However, existing pedagogical approaches vary widely across institutions, regions, and curricular structures, ranging from software-focused instruction to more holistic, design-centered applications. This study presents a comprehensive review of BIM education in architecture by synthesizing trends, pedagogical models, and implementation strategies reported between 2010 and early 2025. A hybrid review design was employed by combining PRISMA-based systematic procedures with scoping and comparative analysis. Bibliometric mapping of 399 BIM education publications identified major research clusters and global trends, while an in-depth analysis of 31 architecture-focused studies revealed seven thematic categories encompassing curriculum integration, design studio pedagogy, immersive technologies, collaborative models, and algorithmic approaches. The findings show a gradual shift from tool-based training toward integrated studio environments where BIM supports design creativity, interdisciplinary coordination, and process-based learning. Persistent challenges—such as balancing technological proficiency with design thinking, adapting faculty expertise, and aligning curricula with industry expectations—continue to hinder deeper integration. Based on the synthesis, this study proposes an integrated educational framework that connects technological competence, design creativity, and collaborative cognition, offering guidance for the next stage of BIM-enabled architectural education. Full article

This study explores the core competencies required for sales and marketing roles in the rapidly evolving NEV sector. Adopting an exploratory sequential mixed-methods design, it employs a big data-driven approach to construct a competency framework: web crawlers collected NEV-related recruitment data across over 20 major Chinese cities, the Latent Dirichlet Allocation (LDA) model identified core competency items, and a multi-dimensional consensus scoring process via the Nominal Group Technique (NGT) refined the framework. The resulting validated model comprises nine thematic clusters, reflecting a shift from internal combustion engine (ICE) vehicle sales’ traditional skill set. Beyond enriching conventional competencies (customer reception, sales service, CRM, sales support), it highlights emerging capabilities: live-streaming/short-video marketing, digital media operations, and ecosystem-oriented resource collaboration. Further, NGT-based multi-dimensional evaluations (frequency, importance, difficulty) generated a four-quadrant matrix, offering actionable guidance for vocational education and corporate training (VET) curriculum design. Theoretically, this study redefines digital-era automotive sales roles: not mere product sellers, but core actors in user experience co-creation and ecological value integration, which enriches discourse on sales role evolution. Full article

Construction 4.0 refers to the integration of automation, artificial intelligence, and the Internet of Things (IoT) in the construction industry, which has changed traditional construction practices. MSc courses play a crucial role in developing the next generation of leaders within the construction industry by equipping graduates of these courses with advanced technical, managerial, and strategic skills, including the arrival of Construction 4.0. As future professionals and construction industry leaders, it is necessary to identify the current level of awareness and understanding of Construction 4.0 amongst master’s level students. As such, this paper studies these areas to help identify the gaps in education and training requirements—essential for matching academic programs with industry needs. Through the use of a survey-based approach with 112 MSc students on various Construction Management courses, a series of revealing results were obtained. The results presented herein indicate that there is a shared definition of what constitutes Construction 4.0 amongst engineering management students. However, while they are relatively aware of Construction 4.0 technologies, they do not differentiate strongly between Industry 4.0 and Construction 4.0. Therein, they are ambivalent as to the role of Education 4.0 in improving this situation. Key to this is the requirement to keep up with industry needs. The lack of application of Construction 4.0 means students lack the necessary ‘practical skills’ to implement innovations on real construction sites. Students advocated for more hands-on training, industry-linked projects, and guest lectures within the curriculum, alongside developing the essential skills of critical thinking and problem-solving. Changes in the curricula are suggested, achievable through readily existing 4.0 Frameworks. Full article

Predicting the sustainability of national educational systems presents a complex, multifaceted issue due to the intricate connections between education and wider societal, economic, healthcare, and technological sectors. Current educational models tend to be rigid, narrow in focus, and insufficiently responsive to these changing external factors. This research seeks to fill this void by framing education as an ecosystem and creating a methodological framework that merges systems thinking with sophisticated data-driven methods. The study’s aim is to outline, quantify, and analyze the relationships among education-related subsystems to guide the creation of an adaptive, sustainability-focused educational ecosystem. A mixed-methods approach was utilized, incorporating qualitative coding, system mapping, and natural language processing techniques (specifically Word2Vec) to uncover relational patterns within a structured literature set. These findings were integrated with quantitative metrics to assess subsystem efficacy and pinpoint leverage points. The investigation centers on five primary systems in the education ecosystem: Business, Economic, Government, Healthcare, and Sustainability. The Word2Vec analysis identified significant conceptual relationships between these systems, while the quantitative evaluation indicated strong performance across curriculum, policy, and healthcare metrics. Conversely, inclusivity and accreditation displayed weaker outcomes, indicating areas that need focused improvement. The results highlight the benefits of merging systems thinking with NLP-driven relational analysis as a methodological innovation in education research. The study offers evidence-based recommendations for prioritizing factors that can boost system efficacy and create beneficial cross-system ripple effects, aiding in the advancement of adaptive and sustainable educational ecosystems. Full article

Climate change education (CCE) is increasingly recognized as a key lever for responding to the climate crisis, yet its implementation in schools often remains fragmented and weakly transformative. This review synthesizes international research on CCE in secondary education, focusing on four interconnected domains: students’ social representations of climate change (SRCC), curricular frameworks, teaching practices and teacher professional development, and emerging pathways towards transformative, justice-oriented CCE. A narrative review of empirical and theoretical studies reveals that students’ SRCC are generally superficial, fragmented and marked by persistent misconceptions, psychological distance and low perceived agency. Curricular frameworks tend to locate climate change mainly within natural sciences, reproduce deficit-based and behaviorist models and leave social, political and ethical dimensions underdeveloped. Teaching practices remain predominantly transmissive and science-centered, while teachers report limited training, time and institutional support, especially for addressing the affective domain and working transdisciplinarily. At the same time, the literature highlights promising directions: calls for an “emergency curriculum” and deeper curricular environmentalization, the potential of socio-scientific issues and complexity-based approaches, narrative and arts-based strategies, school gardens and community projects, and growing attention to emotions, hope and climate justice. Drawing on a narrative and integrative review of empirical and theoretical studies, the article identifies recurrent patterns and gaps in current CCE research and outlines priorities for future inquiry. The review argues that bridging the knowledge–action gap in schools requires aligning curriculum, pedagogy and teacher learning around four key principles—climate justice, collective agency, affective engagement and global perspectives—and outlines implications for policy, practice and research to support more transformative and socially just CCE. Full article

As quantum technologies advance, there is growing international interest in integrating quantum concepts into secondary education. However, most K-12 quantum education studies rely on self-reported data or informal assessments lacking documented validity. This study aimed to address this gap by developing and validating the Quantum Literacy Test (QLt), a standardized instrument designed to objectively assess upper-secondary students’ understanding of foundational quantum concepts, practices, and perspectives. Grounded in the computational thinking (CT) framework, the QLt was piloted with 819 senior secondary school students in Nigeria and underwent a multi-phase validation process, including expert review, factor analysis, item-response modeling, differential item functioning analysis, and concurrent validity. The QLt demonstrated high internal consistency (α = 0.87) and structural validity. Strong concurrent validity was observed with the Computational Thinking Test (r = 0.65), and moderate validity was observed with a Spatial Ability Test (r = 0.32). However, machine learning models explained less than 40% of QLt score variance, suggesting the domain-specific nature of quantum literacy. We recommend future research to expand the QLt across diverse cultural contexts and to increase item coverage of quantum practices and perspectives. The QLt offers a valuable tool for evaluating curriculum effectiveness and monitoring equity in quantum education, thereby contributing to a more inclusive quantum-ready workforce. Full article

(1) Background: Interprofessional supervision is an emerging approach in health professions education that strengthens collaborative practice competencies while maintaining profession-specific expertise. Understanding current evidence regarding supervision models, outcomes, and implementation factors is crucial for advancing this field. (2) Methods: This narrative review analyzed 28 studies, including quantitative, qualitative, mixed-methods studies, and systematic reviews. Studies were analyzed for supervision models, outcome measures, evidence of effectiveness, and implementation factors. (3) Results: Six categories of interprofessional supervision models were identified: clinical practice-based, group supervision, competency-based training, skills training, case-based learning, and mentorship/coaching. Across models, interprofessional supervision consistently enhanced collaborative competencies, professional development, clinical skills, and organizational outcomes. Organizational support, structured curricula, interprofessional leadership, and individual readiness facilitated implementation success. Barriers included limited resources, professional silos, and challenges in curriculum integration. (4) Conclusions: Interprofessional supervision shows consistently positive outcomes across diverse models and settings, though more rigorous research designs and standardized outcome measures are needed. Successful implementation requires systematic attention to multiple factors at multiple levels, from organizational support to individual readiness. Interprofessional supervision is positioned for significant advancement through the application of implementation science frameworks and continued research on optimal model characteristics and implementation strategies. Full article

Traditional Sound Event Detection (SED) approaches are based on either specialized models or these models in combination with general audio embedding extractors. In this article, we propose to reframe SED as an object detection task in the time–frequency plane and introduce a direct adaptation of modern YOLO detectors to audio. To our knowledge, this is among the first works to employ YOLOv8 and YOLOv11 not merely as feature extractors but as end-to-end models that localize and classify sound events on mel-spectrograms. Methodologically, our approach (i) generates mel-spectrograms on the fly from raw audio to streamline the pipeline and enable transfer learning from vision models; (ii) applies curriculum learning that exposes the detector to progressively more complex mixtures, improving robustness to overlaps; and (iii) augments training with synthetic audio constructed under DCASE 2023 guidelines to enrich rare classes and challenging scenarios. Comprehensive experiments compare our YOLO-based framework against strong CRNN and Conformer baselines. In our experiments on the DCASE-style setting, the method achieves competitive detection accuracy relative to CRNN and Conformer baselines, with gains in some overlapping/noisy conditions and shortcomings for several short-duration classes. These results suggest that adapting modern object detectors to audio can be effective in this setting, while broader generalization and encoder-augmented comparisons remain open. Full article