Search Results (1,379)

This study examined rural pre-service teachers’ preferences for online learning platforms, Telegram, WhatsApp, and Moodle discussion forums in the Open Distance e-Learning environment. This group of students experiences digital illiteracy, limited access to assistive technologies, and network challenges, which may prevent them from optimally utilising formal learning platforms such as Moodle. They can, however, use Telegram and WhatsApp, as they regularly engage informally on these platforms. Against this backdrop, this study explored rural pre-service teachers’ experiences with Moodle and these social media platforms in an Open-Distance e-Learning space. This study employed a descriptive, qualitative case study with semi-structured interviews, guided by Siemens’ Connectivism theory. Fifteen student teachers from the College of Education in an ODeL institution were purposively sampled to provide in-depth insights into their lived experiences of platform use. The findings revealed that, although each platform served a unique instructional function, their perceived professionalism, safety, and interactivity differed substantially. Social media platforms such as Telegram and WhatsApp were lauded for their immediacy, accessibility, and low bandwidth usage, chiefly among rural pre-service teachers from economically disadvantaged communities. However, participants perceived these platforms as unprofessional, disruptive, and unsafe. Conversely, Moodle’s discussion forum was viewed as a credible, structured space that fostered academic discipline through the presence and guidance of lecturers. These contrasting perceptions highlight tensions between accessibility and academic regulation within ODeL environments. Although prior studies support incorporating social media platforms into LMSs, this research extends this discourse by emphasising the need to balance accessibility, interaction, and academic integrity within resource-constrained contexts. The study concludes that social media platforms and discussion forums can complement each other in ODeL, encouraging student interaction and inclusion, while discussion forums ensure educational rigour, safety, and institutional integrity. Full article

Three-dimensional (3D) path planning for Unmanned Underwater Vehicles (UUVs) in typical marine operating conditions presents high-dimensional, non-convex optimization challenges due to undulating seabed topography, underwater threat sources, and coupled multi-physical constraints. Existing studies lack multi-strategy collaborative optimization mechanisms specifically designed for UUV 3D marine navigation constraints, thereby hindering the simultaneous achievement of real-time performance, safety, and energy efficiency in path planning. This paper first develops a comprehensive multi-dimensional cost function based on the dynamic characteristics of UUV underwater 3D navigation, operational rules for typical marine operating conditions, and safe navigation requirements through mathematical modeling, thereby formally transforming the UUV 3D path planning problem in typical marine operating conditions into a multi-constrained nonlinear global optimization problem. To address this challenge, an Adaptive Cooperative WMA (ACWMA) is proposed. The key improvements include: (i) an adaptive parameter switching and Lévy flight disturbance mechanism to balance exploration and exploitation capabilities; (ii) an optimal value leadership strategy to accelerate convergence; and (iii) a team collaborative learning mechanism to enhance population optimization efficiency. Algorithm benchmark performance is validated using the CEC 2017 standard test suite, while comparative and ablation experiments are conducted in multi-gradient complex marine 3D scenarios. The statistical significance of the algorithm performance improvement is verified using the Wilcoxon rank-sum test. The proposed ACWMA achieves a significant performance improvement of 8.71% over the suboptimal WMA in terms of core performance metrics and generates low-energy-consumption 3D paths that satisfy multiple constraints. These findings provide valuable engineering insights for 3D path planning in UUV autonomous operations within typical marine operating conditions. Full article

Transfer learning offers a practical way to improve neural machine translation in low-resource settings, but its effectiveness depends on both the choice of transfer language and the amount of target-language data available for adaptation. In this study, we examine these factors specifically for Polish–English translation using mBART. We evaluate Czech, Russian, and German as parent languages and extend the analysis with a combined Slavic parent model trained on Czech and Russian. The models are compared across 0-shot, 10-shot, 100-shot, 1k-shot, and 10k-shot settings. Within this Polish–English mBART setting, Czech provides the strongest zero-shot performance, while Russian and German improve substantially as Polish fine-tuning data increases and achieve the strongest results at higher shot levels. The paper therefore analyzes selected transfer-language configurations rather than a formally measured similarity variable. The results suggest that, in this setup, transfer-language choice matters most when no Polish supervision is available, whereas larger amounts of Polish data can compensate for weaker initial transfer alignment. Full article

Artificial intelligence (AI) systems increasingly depend on complex, multi-stage supply chains that incorporate pre-trained models, third-party datasets, open-source libraries, and automated training pipelines. This dependency creates a rapidly expanding attack surface in which model poisoning, dependency compromise, and provenance manipulation can undermine system integrity long before deployment. Existing AI governance frameworks—including the NIST AI Risk Management Framework and NIST’s Secure Software Development Framework—acknowledge supply chain risks but do not define a verifiable model provenance structure or cryptographically durable integrity guarantees. Simultaneously, the transition to post-quantum cryptography (PQC) introduces new requirements for long-lived AI artifacts: classical digital signatures used to verify model lineage, dataset integrity, and pipeline attestation will become vulnerable to quantum-enabled forgery within the expected operational lifetime of many AI systems. This paper synthesizes evidence from policy, standards, and benchmark sources to characterize the emerging AI supply chain threat landscape and identify cryptographic dependencies that the PQC transition disrupts. We propose a formal Model Bill of Materials with PQC-safe extensions (MBOM-PQC), a unified signing and attestation pipeline integrating ML-DSA and hybrid signature modes, and a five-level Supply Chain Assurance Maturity Model (SCAMM) supporting repeatable organizational evaluation. Together, these contributions aim to provide a structured foundation for AI supply chain integrity, supporting verifiable model lineage, authenticity, and trustworthiness through the PQC transition and beyond. The framework is presented as a design-science contribution comprising three integrated artifacts and is extended with operational guidance for continuous-learning pipelines (§6.5), a formal scoring methodology for organizational assessment (§7.3.5), and a hardware-root-of-trust migration cost matrix (§8.3.6). Full article

Makerspaces can serve as distributed urban infrastructures for repair, reuse, tool sharing, and peer learning, yet their contributions to circular economy (CE) goals often occur without being explicitly recognized or framed as CE practices. Inspired by practice theory and the literature on quiet sustainability, this study introduces implicit circularity as circular practices enacted without an explicit sustainability/CE framing by participants, and examines how such practices shape bottom-up circular transitions. Using reflexive thematic analysis informed by constructivist grounded theory procedures, we examined three linked questions: which circular practices occur in makerspaces and how they cluster into domains, how these practices vary across makerspace types, and which barriers and governance arrangements shape makerspaces’ consolidation as circular urban infrastructure. A qualitative multi-method design was employed in Czechia, combining field mapping with in-depth qualitative inquiry. Data included 40 semi-structured interviews with makerspace founders and operators, documentary analysis based on websites, social media, event listings, rules, and other documents, and 21 observations. Using reflexive thematic analysis informed by constructivist grounded theory procedures, we analyzed how circular practices cluster into domains, how implicit versus explicit circularity varies across makerspace types, which barriers constrain makerspaces’ consolidation as circular urban infrastructure, and what governance arrangements could mitigate them. Circularity was dominated by implicit, routine practices rather than formal, CE-branded programs. Three practice domains were identified: repair and maintenance, material flows, and learning/education. Explicit programming was comparatively less common and context-dependent. Barriers formed a reinforcing system spanning institutional fragmentation and coordination deficits, capability gaps, infrastructural constraints, and tensions around autonomy and legitimacy, which together kept many circular contributions low-visibility. Makerspaces constitute an under-recognized form of circular micro-infrastructure that couples technical capacity with social learning and can translate CE ambitions into everyday practice. To mobilize these latent capacities, cities need hybrid governance, especially light-touch coordination platforms, long-horizon operational support, and integration of makerspaces into municipal material-flow systems and repair/reuse strategies. The study offers a practice-based framework and a cross-case typology to support comparative research and grounded urban CE policy design. Full article

Obstructive sleep apnea (OSA) is a highly prevalent yet frequently underdiagnosed condition that is associated with significant cardiopulmonary, metabolic, and neurocognitive outcomes. Risk factors for OSA overlap with illnesses commonly observed in intensive care unit (ICU) patients, resulting in a disproportionately elevated burden on healthcare. This structured narrative review synthesizes current evidence regarding the prevalence, diagnostic challenges, and clinical implications of obstructive sleep apnea (OSA) in critically ill adults admitted to intensive care units (ICUs) using PubMed, EMBASE, and Scopus. Key search terms included “obstructive sleep apnea,” “ICU,” and “critical illness.” Results showed that OSA is present in up to 60–70% of ICU patients, yet only ~5% are formally diagnosed during hospitalization. Underdiagnosis is linked to prolonged mechanical ventilation, extubation failure rates as high as 30%, 2-fold higher perioperative complication rates, cardiovascular instability, 1.8-fold greater 30-day ICU readmission rates, and 2.2-fold mortality. Standard screening tools have limited applicability in ICU patients. Emerging alternatives, such as overnight oximetry, polygraphy, and machine learning models lack validation. Our analyses reveal that current diagnostic and treatment strategies are poorly adapted to critically ill patients. Integration of OSA as a part of ICU management, diagnosis, and intervention may reduce readmissions and mortality. Full article

The structural integrity and risk management of long-distance oil and gas pipelines are critically challenged by multi-source data heterogeneity, complex multi-physics degradation mechanisms, and the dynamic nature of operational environments. Traditional monolithic artificial intelligence models struggle with cross-domain knowledge fusion and often suffer from historical context forgetting over decades-long infrastructure lifecycles. To address these bottlenecks, this paper proposes an evolutionary digital twin framework driven by a collaborative architecture between small specialized models and a large general model. Specifically, the framework encapsulates physics-informed models (e.g., corrosion prediction and geohazard evaluation) as domain expert agents to guarantee rigorous numerical computation at the edge, keeping sensitive operational data strictly localized. To synthesize conflicting localized risks, a locally deployed, privacy-preserving large language model acts as a central cognitive hub. This hub utilizes external knowledge retrieval and structured reasoning to formulate transparent, multi-objective intervention strategies. Furthermore, a continuous feedback learning mechanism is introduced to capture tacit expert knowledge. By formalizing human operational interventions into historical memory and employing parameter stabilization techniques, the system dynamically updates its knowledge base while effectively mitigating catastrophic forgetting. Ultimately, the proposed framework provides a reliable and privacy-compliant methodology, significantly enhancing the interpretability and predictive foresight of pipeline integrity management. Full article

The rapid advancement of Large Language Models (LLMs) has significantly complicated the distinction between AI-generated and human-written texts. This challenge becomes particularly pronounced in formal and structurally constrained texts, such as academic writing. In this study, a deep learning approach based on the XLM-RoBERTa architecture is proposed for detecting deepfake (DF) texts, with a focus on achieving strong generalization capability within the academic domain. A large-scale dataset comprising 63,000 human-written and AI-generated texts (from Llama-3.1, Gemma-2, Qwen-2.5, Phi-3, Falcon, and Mistral) was constructed. The proposed multi-model data strategy is designed to encourage the model to learn structural and stylistic distinctions between human and AI-generated texts, rather than memorizing model-specific stylistic patterns, thereby reducing false positive rates, particularly for formal human-written content. To analyze the model’s learning behavior, no preprocessing was applied to the training data. The model was evaluated on two independent test sets (preprocessed and non-preprocessed), neither of which was seen during training. Experimental results show that the model achieves an F1-score of 99.76% on the validation set, while maintaining 93.42% accuracy and 94.67% recall on the preprocessed (Zero-Artifact) test set. These findings indicate that the model relies on inherent linguistic and structural patterns of AI-generated text in formal context, rather than dataset-specific superficial artifacts, suggesting improved robustness and generalization in academic integrity applications. Full article

The ability to identify emotions based on written text is one of the core areas of Natural Language Processing (NLP) and has many applications in areas such as mental health monitoring, sentiment analysis, and dialogue systems. This study proposes an explainable emotion recognition (EER) framework built on a fine-tuned RoBERTa-base model trained on the Emotions for NLP dataset with an accuracy of 92.4% and a weighted F1 score of 92.5%. To interpret the decision process of the EER model, we systematically applied four complementary explainable artificial intelligence (XAI) techniques to provide explanations and insights into how the model makes its predictions: SHAP for global token-level feature attribution, LIME for local instance-level explanations, multi-head attention visualization for structural interpretability, and integrated gradients via Captum for axiom-satisfying gradient-based attribution. Each of these four methods provides complementary multi-perspective views of EER model behavior, which can help increase model transparency, identify potential biases, and enable the responsible use of transformer-based models in critical environments (e.g., those requiring formal clinical documentation). Our experiments consistently show that the EER model identifies tokens as having the highest emotional expression level as the strongest predictive feature across methodological perspectives, with strong evidence of cross-methodological agreement regarding the semantic coherence of learned representations. Our findings have direct implications for the responsible implementation of AI-based emotion recognition systems in mental health support systems, where model user-interface transparency, bias mitigation, and clinical trust are necessary to ensure quality patient care. Full article

Prosumer communities, aggregations of residential and commercial entities equipped with distributed energy resources (DER), including photovoltaic systems, battery storage, and flexible loads, are emerging as critical organizational units in decarbonising smart grid architectures. Managing these communities effectively requires balancing economic efficiency with equity, autonomy, and environmental sustainability, objectives that conventional centralized control methods and existing multi-agent reinforcement learning (MARL) implementations fail to address simultaneously. This article proposes a value-aligned hierarchical multi-agent reinforcement learning (VA-HMARL) framework as a formally unified architecture that embeds equity (Jain’s Fairness Index J ≥ 0.90), individual autonomy, and carbon sustainability as hard constraints within the MARL reward structure. The framework integrates: a multi-objective Value Alignment Module (VAM) combining economic, fairness, sustainability, and comfort objectives; attention-based implicit coordination for scalable agent interaction; and differentially private federated policy aggregation (ε = 1.0, δ = 10⁻⁵) for GDPR-compliant collaborative learning. Simulation on a 20-prosumer community modelled on the IEEE 33-bus feeder over 10 Monte Carlo runs (300 episodes each) demonstrates: a 6.2% energy cost reduction versus the Rule-Based baseline (p = 0.0004); a Jain’s Fairness Index of 0.912 ± 0.031 at policy convergence (final 50 episodes), satisfying the J ≥ 0.90 community equity floor; and an 18.0% reduction in CO₂ emissions. The economic efficiency trade-off relative to performance-optimized MARL baselines is limited to 2.4%, within the 5% design target. These results establish VA-HMARL as a technically feasible and ethically grounded paradigm for autonomous decentralized energy governance. Full article

The rapid proliferation of Internet of Things (IoT) devices in cloud-based e-learning platforms has posed significant security risks, particularly in protecting learner information, authentication of devices, and safe communication in the highly heterogeneous learning settings. Current cryptographic solutions are largely based on classical public-key infrastructure (PKI) protocols such as RSA and ECC, which will become vulnerable with the advent of large-scale quantum computers capable of executing Shor’s algorithm. In addition, traditional perimeter-based security models are inadequate for handling the dynamics, scattered, and resource-limited characteristics of IoT-enabled educational systems. As a solution to these problems, this paper introduces ZeroTrustEdu, a scalable zero-trust cryptographic solution that combines lightweight post-quantum key management with adaptive trust scoring of cloud-connected IoT e-learning infrastructure. The proposed framework makes three fundamental contributions namely: (1) a hierarchical zero-trust security model with no implicit trust, operating across device, edge, and cloud layers; (2) a lightweight key distribution protocol based on the Module-Lattice Key Encapsulation Mechanism (ML-KEM) compliant with NIST FIPS 203 standards and (3) an adaptive behavioral trust scoring engine that dynamically adjusts device and user trust levels based on real-time interaction analytics. The architecture is evaluated using extensive NS-3 network simulations with up to 100,000 concurrent IoT nodes with formal security analysis under Chosen Plaintext Attack (CPA) and Chosen Ciphertext Attack (CCA) threat models. Comparative evaluation against RSA-2048, ECC-P256, and AES-256 baselines demonstrates that, ZeroTrustEdu delivers a 62% ± 3% (95% CI, 10 independent runs) reduction in ML-KEM encapsulation latency (12.8 ms for key encapsulation/decapsulation, contributing to a complete device authentication latency of 47.3 ms including ML-DSA signature operations), 45% reduced communication overheads, and 38% reduction in energy consumption on ARM Cortex-M4 constrained devices compared to RSA-2048 and achieves provable post-quantum security reducible to the hardness of the Module Learning With Errors (MLWE) problem. These findings demonstrate that the proposed architecture provides a viable, scalable, and quantum-resilient security solution for next-generation IoT-enabled e-learning environments. The cryptographic security of ZeroTrustEdu is guaranteed at the primitive level through NIST-standardized ML-KEM (FIPS 203) and ML-DSA (FIPS 204), with IND-CCA2 and EUF-CMA security formally proven in the respective standards; full protocol-level formal verification using automated theorem provers (ProVerif, Tamarin) is identified as valuable future work to rule out protocol-composition vulnerabilities beyond primitive-level guarantees. Full article

Credit risk evaluation is a key factor in financial intermediation, regulatory capital provision, and risk management in the portfolio. In this study, we compare the deep learning performance for probability-of-default (PD) estimation with a structured financial econometric model using loan-level data of an Indian non-banking financial agency between May and August 2025. Using the interpretation of PD as a conditional expectation, which is in line with reduced-form default-intensity models, we compare deep learning, logistic regression, and gradient boosting using a pure time-based out-of-sample design. Model assessment focuses on discrimination and calibration, where the area under the precision–recall curve (AUC-PR), Brier score, log-loss, and Hosmer–Lemeshow goodness-of-fit tests are utilized. The findings show that deep learning achieves higher accuracy in terms of calibration but a lower Brier score by about 18; this gap could be reduced by comparing logistic regression with statistically significant improvements in formal tests that compare forecasts. In portfolio back-testing, better probability scaling is translated into an actual loss reduction of about 12–13% for the August 2025 cohort. Although the improvements compared with the advanced ensemble techniques are moderate, the results indicate that deep learning improves the estimation of conditional default probabilities because of the better nonlinear modeling and upper-tail risk perception. This study contributes to the literature via its incorporation of machine learning and credit risk assessment into a formalized risk management and econometric assessment system. Full article

Implementation remains a central challenge in urban policy, yet the knowledge formats designed to bridge the gap between policy goals and on-the-ground delivery remain under-examined. This study treats 250 UN-Habitat Best Practice reports not as proof of effectiveness but as a standardized genre through which local interventions are narrated, compressed, and made portable for replication. We extract three focal sections, namely Results, Lessons Learned, and Transferability, apply systematic thematic coding with 906 open codes consolidated into axial categories, and compute co-occurrence networks using Jaccard similarity and Lift to detect thematic bundles, holes, and silos within and across sections. Three findings emerge. First, the reporting repertoire narrows progressively, as mean thematic richness declines by 28.2% from Results to Transfers while concentration increases 4.2 times, with substantive dimensions such as governance, equity, sustainability, and evidence losing prevalence to circulation-oriented themes. Second, formal bundle detection yields zero qualifying pairs across all six matrices, indicating a loosely coupled reporting grammar anchored by generic silos rather than integrated implementation packages. Third, structural holes concentrate at the pipeline’s end, where infrastructure transfer and sustainability as transferable value are the most systematically disconnected themes. These patterns reveal a portability paradox in which the reporting format achieves institutional legibility, making practices comparable within a shared vocabulary, but progressively filters out the physical, evidentiary, and context-sensitive content that operational reproduction would require. Full article

We designed a supervised machine learning framework to predict standard Gibbs free energies, ΔG°, of formal hydrogen atom transfer (f-HAT) for phenolic antioxidants across different radicals and media, enabling rapid and chemically interpretable screening. We curated a DFT dataset of 71 molecules (phenolic compounds and anthocyanidins), with 207 reaction sites, 10 radical reactive oxygen/sulfur species, and three environments (leading to a total of 6210 ΔG° values). The models amass 106 numerical RDKit descriptors, augmented with one-hot encodings of medium, site, radical, and structural class, and were evaluated through a leave-one-molecule-out protocol. Among the tested regression algorithms, the random forest regressor provides the best balance of accuracy and robustness with both R² test (≈0.94) and MAE (2.74 kcal mol⁻¹; RMSE (≈5.0 kcal mol⁻¹)), close to DFT chemical accuracy. The feature-importance analysis revealed that “electronic” and “experimental” (site/group) descriptors primarily drive predictions, with the radical’s maximum absolute partial charge being the most important descriptor in the prediction of a radical’s ΔG°. These results suggest that descriptor-driven RF (Random Forest) models can generalize across chemical space to provide interpretable ΔG° predictions, providing a path for chemists towards a scalable route to prioritize antioxidant candidates for broader molecular families. Full article

Background/Objectives: Dental metallic implants cause severe streaking artifacts in kilovoltage CT (kVCT), compromising dose calculation in radiotherapy (RT) treatment planning. The purpose of this study is to assess the dosimetric agreement of synthetic MVCT (sMVCT) images generated from artifact-affected kVCT using a deep learning network with respect to true MVCT (tMVCT) acquired at the treatment machine. Methods: Nineteen head and neck cancer patients with dental metallic implants treated with RT were included. Planning kVCT images were converted to sMVCT using Metal Artifact Reduction through Domain Transformation Network (MAR-DTN), a UNet-inspired deep learning network. The sMVCT images were rigidly registered to true MVCT (tMVCT) acquired on the Hi-Art II Tomotherapy system. Mean Hounsfield Unit (HU) values were compared across seven structures (thyroid, bilateral parotids, brainstem, spinal cord, GTV, PTV70) using pairwise Wilcoxon tests and Two One-Sided Tests (TOST) for statistical equivalence within a pre-specified margin of ±20 HU (corresponding to a 2% deviation in physical density). Dose distributions were recalculated on sMVCT using the AAA algorithm and compared to reference tMVCT-based plans via dose–volume histogram (DVH) metrics, evaluated for equivalence by TOST within a margin of ±2% of the prescribed dose (±142 cGy of 70.95 Gy), and via 3D gamma index, evaluated by one-sided non-inferiority test against the clinically accepted thresholds of 90% (2 mm/2%) and 95% (3 mm/3%). A pre-specified sensitivity analysis was performed by repeating all comparisons on the strictly independent sub-cohort (n = 16) excluding three patients drawn from the MAR-DTN training set. Results: All seven anatomical structures showed statistical equivalence between sMVCT and tMVCT under the ±20 HU margin (TOST p < 0.05; mean HU differences in the range −1.1 to +8.4 HU; all Wilcoxon p > 0.05). All nine DVH metrics achieved formal dosimetric equivalence within ±2% of the prescribed dose (TOST p < 0.05). Mean 3D gamma pass rates were 94.3% (95% CI: 89.3–97.1) for the 2 mm/2% criterion and 97.6% (95% CI: 94.8–99.0) for the 3 mm/3% criterion, both formally non-inferior to the respective clinical thresholds (p < 0.0001). Residual gamma failures were concentrated at the patient surface, consistent with inter-session repositioning uncertainty rather than errors in synthetic image generation. Sensitivity analysis on the n = 16 sub-cohort confirmed all conclusions, with mean HU and DVH differences smaller than in the full cohort for the structures showing the largest mean differences, and comparable for the remaining structures, with all TOST equivalence and gamma non-inferiority tests confirmed in both cohorts. Conclusions: sMVCT images generated via MAR-DTN show dosimetric agreement with physically acquired tMVCT in head and neck patients with dental implants, formally demonstrated by TOST equivalence within ±2% of prescribed dose for all DVH metrics. The combined HU and gamma index framework presented here represents a promising quality assurance approach for AI-based synthetic imaging tools in radiotherapy, pending validation in larger prospective multicentre cohorts. Full article