Search Results (3,893)

The increasing integration of artificial intelligence (AI) into early childhood education has led to ethical concerns regarding children’s privacy, fairness, and developmental appropriateness. While ethical knowledge is recognised as an important component of teachers’ professional knowledge, existing instruments such as the Technological Pedagogical Content Ethical Knowledge (TPCEK) framework were developed for preservice teachers in general and do not fully capture the ethical issues specific to AI-supported preschool education. In this study, we adapted and validated a Self-Perceived Ethical Knowledge Scale for Chinese preservice preschool teachers based on the eight ethics-related dimensions of TPCEK. A two-phase sequential design was employed. In Phase 1, a three-round modified Delphi study was conducted, involving 20 experts who refined an initial 40-item pool into a 33-item scale. In Phase 2, we examined the scale’s psychometric properties by performing an exploratory factor analysis (EFA) on a pilot sample (n = 289)—with factor retention corroborated by principal-axis parallel analysis—and confirmatory factor analysis (CFA) on a separate main-study sample (n = 583). The final 31-item, eight-factor scale demonstrated satisfactory model fit (CFI = 0.943, TLI = 0.932, RMSEA = 0.047, SRMR = 0.048), internal consistency (Cronbach’s α = 0.776–0.887; CR = 0.784–0.891), convergent validity (AVE = 0.549–0.622), and discriminant validity (Fornell–Larcker criterion and HTMT < 0.85). A second-order CFA also showed acceptable fit, although it fitted the data significantly less well than the correlated first-order model. The validated scale provides a context-specific instrument for assessing self-perceived ethical knowledge among preservice preschool teachers in AI-enhanced teacher education. Full article

This study examines whether perceived enterprise use of agentic AI capabilities is associated with strategic business development through decision-making efficiency. Cross-sectional survey data were obtained from 318 managers, analysts, and professionals working in enterprises of varying sizes and industries. The model was estimated using partial least squares structural equation modeling. Perceived agentic AI use was positively associated with decision-making efficiency (β = 0.624, p < 0.001) and strategic business development (β = 0.283, p < 0.001). Decision-making efficiency was also positively associated with strategic business development (β = 0.478, p < 0.001). The indirect association was significant (β = 0.298, p < 0.001; 95% CI [0.215, 0.387]), indicating partial mediation. By focusing on goal-directed, context-adaptive, and action-oriented capabilities, the study extends research that treats AI as a broad organizational capability and identifies decision-making efficiency as a plausible process mechanism. Because the data are cross-sectional and self-reported, the findings do not establish causality, verify a specific technical implementation, or demonstrate superiority over conventional or generative AI. Potential relevance to SDGs 8, 9, and 12 is discussed as a conceptual implication of responsible implementation; the survey did not measure SDG outcomes directly. The study recommends that enterprises integrate agentic AI into core decision-making functions through clear governance, workflow redesign, and sustainability-aligned strategic objectives. Full article

Modern network infrastructures are under attack from increasingly sophisticated attacks that static, rule-based defenses cannot adequately mitigate. We introduce a multilayer anomaly detection and mitigation framework for Software-Defined Networking (SDN) environments consisting of four integrated subsystems: (i) a Micro-segmentation Integrated Management and Defense System (MIMDS), (ii) an Adaptive CNN-LSTM-Attention Deep Packet Inspection (MCLA-DPI) module, (iii) a Hybrid Adaptive Cyberattack Prediction (KBGM) framework, and (iv) an AI-driven log analysis pipeline. Detection, prediction and containment operate in parallel (as opposed to conventional approaches) and correlate results through a common risk-scoring mechanism to coordinate policy enforcement via OpenFlow and P4-compatible data planes. The experimental evaluation shows promising performance. MIMDS achieves 94.3%. detection accuracy with full traffic isolation in 10.1 s. MCLA-DPI achieves 98.9% classification accuracy with 14 ms inference latency, outperforming baseline models SVM and LSTM on encrypted traffic. KBGM achieves 98.4% detection accuracy with 7.2 ms mean response time on the CICIDS2017 dataset. All modules are trained in federated learning to preserve data locality with continuous improvements of the global model. The results collectively demonstrate quantifiable improvements over single-paradigm approaches in detection fidelity, response latency, resource efficiency, and privacy compliance. An ablation study isolates the contribution of integration itself. Removing the coordination layer while retaining all four detectors reduces accuracy from 0.892 to 0.634 and raises the false-positive rate from 0.031 to 0.436, while peak rule installation rises from 22.3 to 98.4 rules per second and oscillation events increase by two orders of magnitude; the integrated framework also exceeds its strongest individual subsystem, which reaches 0.831 accuracy at a false-positive rate of 0.117. These figures are obtained from the released reference implementation over a synthetic campaign and are reported separately from the component measurements. Full article

Artificial intelligence (AI) is rapidly transforming employment relationships through automation, algorithmic management, automated decision-making, and workplace surveillance. While these technologies offer opportunities for productivity gains and organizational efficiency, they also generate significant challenges for labour law by reshaping traditional forms of managerial control and creating new risks for workers’ fundamental rights. This article examines how AI is redefining work and labour rights in Peru, with particular attention to equality and non-discrimination, privacy, human dignity, job security, freedom of association, collective bargaining and worker participation, and occupational well-being. Using a doctrinal and comparative legal methodology, the study analyses the Peruvian regulatory framework alongside recent European and international labour standards, including the General Data Protection Regulation (GDPR), the European Union Artificial Intelligence Act, the Platform Work Directive, and ILO Convention No. 193 concerning decent work in the platform economy. The findings indicate that the principal challenge is not automation itself but the increasing use of algorithmic systems to govern employment relationships through opaque and data-driven decision-making processes. Although Peru has adopted Law No. 31814 and its implementing regulation, which establish general AI-governance rules and classify certain employment uses as high-risk, significant employment-specific procedural gaps remain regarding algorithmic management, contestation, evidentiary access, collective participation, and coordinated labour enforcement. The article argues that future reforms should pursue selective regulatory adaptation through a worker-centred approach combining transparency obligations, meaningful human oversight, targeted algorithmic audits and impact assessments, collective participation mechanisms, and effective enforcement. Ultimately, the study concludes that the transition from algorithm to law requires labour-specific procedural safeguards capable of ensuring that technological innovation remains compatible with fundamental rights, decent work, and social justice in the digital age. Full article

AI writing tools like ChatGPT, QuillBot, and Grammarly have reshaped university writing workflows. Controlled studies show gains in grammatical accuracy, student engagement, and self-efficacy, but none of the intervention studies retained in this review directly measured critical thinking or other higher-order outcomes, leaving concerns about overreliance largely untested rather than resolved. Synthesising 68 references published between 2008 and 2025, including 42 empirical studies, this narrative review examines how students utilise generative AI, its influence on core learning outcomes, and necessary institutional adaptations for policy and assessment. To address these shifts, the paper synthesises empirical findings across diverse study designs and proposes a seven-dimension AI literacy framework alongside practical design principles for effectively integrating AI tools into modern writing instruction. Full article

The “lock and key” model of molecular recognition, anchored by the canonical sequence–structure–function paradigm, has successfully served as a basis for designing biomimetic biomaterials for decades. This classical concept associates the function of molecules with a stable three-dimensional structure that recognizes a specific complementary surface. However, biological systems routinely achieve specificity, adaptability, and multifunctionality through protein domains that never adopt a single stable fold. Intrinsically Disordered Proteins (IDPs) and Intrinsically Disordered Regions (IDRs) represent an underutilized functional repertoire in which conformational plasticity, short linear motifs, and phase-separation capacity underpin behaviors that rigid-fold proteins cannot replicate. This Perspective Article extends the molecular biomimetic design framework to the broader range of IDPs/IDRs roles—as effectors triggering downstream signaling, as motion-directing elements interacting with external partners, and as molecular assemblers that organize dynamic complexes while largely avoiding the steric constraints that structured proteins impose. Sequence plasticity, which is greater in disordered domains than in structured proteins, further expands this versatility. Metamorphic and moonlighting proteins further extend the sequence-to-function landscape by encoding multiple folds or functions within a single chain. We discuss how this bio-informed framework guides peptide-based biomaterials design by exploiting the dynamic, multivalent interactions of IDPs. Artificial Intelligence (AI)-guided approaches and machine learning strategies—from supervised classifiers to reinforcement learning loops—that are integrated with experimental feedback to navigate within these high-dimensional design landscapes of intrinsically disordered bio-informed materials systems are provided. Next-generation bio-informed material design must meet challenges driven by clinical, environmental, and industrial needs. Expanding protein matrix in bio-informed material design may enable mimicking complex, dynamic and hierarchical biological functions and materials that are found in Nature. Embracing disorder may even expand functions beyond addressing current challenges. Moving beyond the single-structure paradigm that has dominated the biomimetic field, embracing disorder in biological protein repertoire may give rise to emerging functions in bio-informed biomaterials. These functions are truly adaptive, modular, resourceful, responsive, and sustainable. Full article

Port State Control (PSC) inspections are essential for maritime safety, but limited inspection resources make efficient vessel pre-screening increasingly important. This study investigates whether external vessel images can provide complementary information on visible ship deficiencies for PSC-oriented decision support. Five convolutional neural networks (ResNet18, GoogLeNet, MobileNetV2, DenseNet201 and SqueezeNet) were evaluated, followed by stacking and a hybrid framework combining ResNet18 deep features with handcrafted descriptors, mRMR feature selection, Principal Component Analysis, boosting classifiers and adaptive threshold optimisation. To make the comparison directly reproducible, duplicate image routes were removed, and all model families were evaluated on fixed target-specific 65/20/15 partitions. On the common independent tests, DenseNet201 provided the strongest balanced result for oxidation (balanced accuracy, 0.672; AUC, 0.742), whereas MobileNetV2 produced the largest point estimates for paint deterioration (balanced accuracy, 0.561; AUC, 0.659) and structural corrosion (balanced accuracy, 0.777; AUC, 0.885). Structural-corrosion recall was 3/4 = 0.750 (95% exact CI, 0.194–0.994) for MobileNetV2 and 2/4 = 0.500 (0.068–0.932) for the strict Hybrid PSC model, illustrating the uncertainty associated with rare positive cases. A controlled ablation further showed that changing only the operating threshold increased recall from 0.10 to 0.85 for oxidation and from 0 to 0.70 for paint deterioration, while increasing the alert burden. The results do not establish universal superiority of the hybrid representation; they show that representation and operating point should be interpreted jointly when visual AI is used as complementary PSC pre-screening support. Full article

Generative artificial intelligence (GenAI) is being integrated rapidly across higher education, yet its educational value depends less on adoption intensity than on the institutional conditions through which its use is organized and enacted. This conceptual study develops a six-axis framework for examining institutional articulation across governance, curriculum, assessment, faculty development, digital ethics, and equity. Drawing on a selective integrative analysis and conceptual synthesis, the framework moves from a dimensional taxonomy of institutional interventions to a configurational account of their interdependence. It distinguishes institutional articulation, the process through which institutional functions are connected, interpreted, resourced, and revised, from institutional coherence, the variable degree of alignment that results. Human capability formation provides the substantive educational criterion for evaluating that coherence, distinguishing capability-enhancing configurations from arrangements oriented primarily toward efficiency, automation, or output production. The framework further situates the six institutional axes within a broader relational architecture in which affect, trust, and perceived legitimacy condition enactment; assessment validity and demonstrable learning provide evidence of capability formation; and credential credibility, professional adaptability, stakeholder trust, reputation, and legitimacy constitute potential system-level consequences. The framework is operationalized through phased institutional transformation and illustrative process, outcome, and alignment evidence, and generates four propositions for future empirical research. Full article

The increasing deployment of Renewable Energy Sources (RESs), particularly wind and solar power, plays a critical role in reducing carbon emissions and supporting sustainable energy transitions. However, the large-scale integration of RESs into smart grids introduces significant technical challenges related to frequency stability, voltage regulation, rotor angle stability, power quality, inertia reduction, harmonic distortion, reverse power flow, Sub-Synchronous Interactions (SSIs), and protection coordination. Although numerous review studies have examined renewable energy integration, most focus on high-level frameworks, bibliometric analyses, optimization techniques, or isolated applications of artificial intelligence (AI) while lacking a comprehensive synthesis that bridges AI-driven solutions with the physical dynamics, control mechanisms, and protection requirements of highly renewable power systems. To address this gap, this review provides a comprehensive technical assessment of wind generator topologies, solar inverter architectures, grid-forming and grid-following control strategies, virtual inertia and virtual Synchronous Generator (SG) technologies, adaptive load-frequency control, energy storage integration, protection coordination, and real-time stability enhancement techniques for high-RES smart grids. Furthermore, the review systematically examines the role of AI in frequency regulation, voltage control, harmonic mitigation, predictive operation, parameter optimization, and system resilience. Unlike previous reviews, this study integrates physical-layer perspectives by connecting AI-driven decision-making with practical grid control mechanisms, inverter dynamics, wide-area monitoring, microgrid operation, High Voltage Direct Current (HVDC) interconnections, EV/Vehicle-to-Grid (V2G) integration, and multi-resource energy management. The review identifies key research priorities, including the development of real-time AI-assisted frequency control, adaptive protection schemes for low-inertia systems, coordinated grid-forming inverter control, resilient autonomous grid operation, and scalable multi-energy management frameworks. The findings provide actionable guidance for researchers, utilities, policymakers, and industry stakeholders seeking to enhance stability, reliability, and operational flexibility in future smart grids with very highly renewable energy penetration. Full article

Sustainable agricultural management faces growing challenges from climate change, water scarcity, energy constraints, soil degradation, and the need to ensure both food production and nutritional security. This study proposes an integrated computational framework combining a Global Agricultural Foundation Representation Model (GAFRM), Self-Evolving Explainable Digital Twin (SEDT), Self-Evolving Evolutionary Foundation Optimizer (SEEFO), and an enhanced Green Regenerative Agriculture Sustainability Index (GRASI). The framework operationalizes the food–water–energy–carbon–nutrition (FWEC-N) nexus by explicitly incorporating crop micronutrient density into the agricultural decision architecture and modeling its relationship with regenerative practices such as cover cropping, biochar application, and zero tillage. Using multi-source global datasets, GAFRM learns transferable agricultural representations, SEDT enables adaptive prediction under climate uncertainty, and SEEFO performs five-objective optimization of agricultural productivity, irrigation water use, energy demand, net carbon balance, and overall sustainability, while nutritional quality is evaluated through the MODI outcome indicator. The enhanced GRASI further evaluates nutrient output, soil restoration, carbon storage, and climate resilience within a unified sustainability framework. The framework was evaluated using a global agricultural dataset covering approximately 60 representative countries across six continents and 15 climate zones over the 2000–2026 period. SEDT achieved an RMSE of 3.18, MAE of 2.29, R² of 0.972, and NSE of 0.968, while SEEFO achieved the highest Hypervolume (0.956) and the lowest GD (0.028), IGD (0.039), and Spread (0.162) among the benchmark optimization algorithms. The observed performance differences were statistically significant according to the Wilcoxon signed-rank and Friedman tests (p < 0.05). The findings indicate that integrating nutritional quality with resource efficiency, carbon balance, soil regeneration, and climate resilience provides a more comprehensive basis for evaluating regenerative agricultural strategies. The architecture establishes a fully transparent, explainable decision-support environment through explainable AI (XAI) feature attributions, bridging the gap between digital precision farming, regenerative ecosystem restoration, and sustainable human nutrition under increasing environmental uncertainty. Full article

Artificial Intelligence (AI)-based educational systems increasingly support personalised learning, adaptive feedback, and learning analytics. However, these capabilities are often addressed separately, with comparatively less attention paid to their integration into interpretable, teacher-facing decision-support frameworks. This study presents the e-Teacher Assistant, a human-centred educational framework that integrates xAPI-style Learning Record Store (LRS) analytics, dynamic learner modelling, Sugeno-type fuzzy inference using interpretable IF–THEN rules, adaptive learning support, and a dual Student Model–Teacher Model architecture. The framework was evaluated during a three-month authentic deployment in a vocational education Computer Networks course involving 117 learners and four educators. Learner evaluation combined a structured questionnaire, open-ended responses, and LRS-based behavioural analytics, while educators evaluated the Teacher Model through a questionnaire and qualitative responses. Learners reported predominantly positive perceptions of the system, including ease of use (92.3%), usefulness of feedback (95.7%), helpful interaction with the AI Assistant (94.9%), support for independent learning (88.0%), and support for educators through learning analytics (87.2–88.0%). Fairness and objectivity were also evaluated positively by 86.3% of learners. All four educators evaluated the Teacher Model positively, and all strongly agreed on its overall usefulness and its support for progress monitoring. Prior familiarity with digital learning tools was significantly associated with evaluations across all seven questionnaire dimensions. LRS analytics further documented learner engagement, repeated assessment activity, and contextual AI use; AI-assisted examination support was recorded in 64.42% of learner–chapter records. Overall, the findings provide initial empirical support for the feasibility of integrating interpretable learner modelling, adaptive AI-assisted support, LRS-based learning analytics, and teacher-facing decision support in an authentic vocational education setting while preserving educator oversight and pedagogical responsibility. Full article

Cloud–edge continuums are driving the shift of cloud-native applications from centralized data centers to latency-, mobility-, privacy-, and energy-saving applications. The serverless architecture provides an attractive “Function-as-a-Service” (FaaS) model in this transition, as it is driven by events, elastic and fine-grained, and controlled by the platform. But introducing a mix of heterogeneous edge nodes, fog/MEC resources, regional clouds, and hyperscale data centers creates a seemingly simple FaaS deployment problem to solve with a set of multi-objective orchestration challenges: runtime selection, autoscaling, cold start mitigation, placement, migration, workflow coordination, state management, trust, cost, energy, and carbon. In this article, we provide an extensive critical review of serverless functions in cloud–edge environments. While some surveys are narrowly focused on aspects of autoscaling, offloading, IoT, or security, the review brings together architectural evolution, runtime mechanisms, platform ecosystems, governance issues, sustainability issues, and emerging applications using AI. It builds a multidimensional taxonomy ranging from runtime systems, autoscaling, cold start mitigation, function placement, and offloading/migration, to workflow orchestration, state and data management, intelligent scheduling, security, sustainability, and industrial serverless platforms. It also presents a built-in conceptual model that connects application needs, runtime environment, orchestration intelligence, governance policies, and system-level results. The synthesis reveals that cloud–edge serverless systems need accountable placement, state-aware workflows, reproducible benchmarking, trustworthy orchestration, and carbon-aware lifecycle control, which can be achieved only by going beyond latency and elasticity. The paper ends with research directions on adaptive, interoperable, explainable, and sustainable serverless systems on the cloud–edge continuum. Full article

Wax deposition in pipelines is a major issue in the oil and gas industry. Estimating the main characteristics of the wax deposits is crucial in mitigating their negative impact. Thus, developing predictive models plays an important role in managing wax deposition. Artificial intelligence (AI)-based models have proven their effectiveness and accuracy, along with several advantages such as ease of use, flexibility, and adaptability. This paper presents a comprehensive review of 41 primary studies reporting over 100 individual AI-based models used for wax deposition prediction, including support vector machines (SVMs), feedforward neural networks (multilayer perceptron, RBFNN, cascade-forward, and others), neuro-fuzzy systems, and tree-based models, together with hybrid and metaheuristic-optimized variants. In addition, AI-based models that integrate multiple single-model predictors or optimization techniques were also reviewed. This paper discusses the underlying principles, applications, strengths, and limitations of these AI-based prediction techniques and concludes with an outlook on future research directions in AI-driven wax deposition prediction. According to the reviewed papers, and by analyzing the errors reported, AI-based models can successfully predict the wax deposition rate, deposited weight, thickness, wax appearance temperature (WAT), and wax disappearance temperature (WDT). AI-based models have high potential to compete with conventional models and efficiently contribute to wax deposition control and management. This review finds that although these models routinely report high accuracy (R² > 0.95), such results are typically obtained on small, frequently reused datasets with limited validation. While gradient-boosting tree ensembles are the most frequent winners in recent head-to-head comparisons, no single model family is consistently superior across prediction targets. Full article

Agentic AI can interpret information, plan, make workflow decisions, and use enterprise tools. Yet technical capability does not establish authoritative meaning, legitimate process state, organisational permission, or accountable execution. The challenge is to preserve adaptability while ensuring that consequential actions remain governed. This article develops a domain-independent conceptual framework for governed agentic AI in enterprise workflows based on bounded autonomy, delegated authority, and controlled execution. A consequential action or workflow decision selected by an agent is treated as a proposed action. It may change enterprise state only after independent controls confirm semantic validity, procedural admissibility, policy compliance, and delegated authority. Actions that pass these controls and remain within a task envelope may proceed automatically through controlled enterprise tools. Those exceeding thresholds for consequence, irreversibility, uncertainty, data sensitivity, value, or organisational policy are escalated to an accountable human. Human oversight is therefore risk-proportionate rather than required for every action. The framework integrates enterprise ontologies, governed knowledge graphs, Business Process Model and Notation (BPMN) orchestration, policy and decision services, controlled tool execution, and provenance within explicit responsibility and authority boundaries. A review-informed design-science process synthesises evidence into five connected control gaps and derives ten design requirements, operationalised through task envelopes, capability and authority relations, lifecycle states, exception paths, and conformance criteria. An illustrative online-shopping order exception demonstrates the control logic, while a control-loop walkthrough and ten failure and adversarial conditions trace requirements-to-control mappings, responsibility separation, and defined recovery paths. The framework provides a systematic basis for governing agentic AI as an adaptable enterprise participant. It supports risk-proportionate autonomy, auditability, accountability, and regulatory evidence. The analytical walkthrough supports conceptual coherence and design plausibility but does not establish deployed effectiveness or legal compliance. Full article

Harmful algal blooms (HABs) are escalating global threats to aquatic ecosystems, water security, and public health. Although conventional monitoring and management approaches, ranging from in situ sensing to predictive ecological models, have advanced bloom detection and risk assessment, their effectiveness is often constrained by data scarcity, transferability, and real-time applicability. Artificial intelligence (AI) offers transformative capabilities across the HAB management continuum, from detection to decision support, positioning AI as a cornerstone of next-generation strategies to mitigate bloom risks. This review synthesizes recent progress in AI applications, including automated phytoplankton identification, remote sensing analysis, predictive modeling, and decision-support systems, and evaluates classical machine learning, deep learning, and automated machine learning (AutoML) approaches. This review highlights how integrating AI with conventional ecological knowledge and expert judgment can yield adaptive, scalable hybrid intelligence frameworks. By integrating technological innovation with established monitoring practices, next-generation HAB management can shift from reactive responses to proactive strategies. Full article