Search Results (5,543)

The aim of this study was to design and validate a questionnaire to assess students’ perceptions of the use of Artificial Intelligence (AI) tools in academic tasks (PEHIA-TA). To determine the psychometric properties of the PEHIA-TA, a descriptive, exploratory and confirmatory factor analysis was carried out. The sample used in this study consisted of 546 students. The results confirmed that it is a valid and reliable scale with a five-factor structure: “Uses of Artificial Intelligence (AI)” (student opinion, knowledge and experience in relation to AI); “Perceptions of skills needed to use AI” (type of skills they consider necessary to work with this type of tool); “Plagiarism and lack of academic integrity” (issues related to what the student considers plagiarism and lack of academic integrity in order to identify possible risks or associated moral dilemmas); “Perception of the benefits of AI” (assessment of the beneficial aspects of the use of AI in the academic context by students); and “Perception of the problems of AI” (analyses how students assess the problems associated with the use of AI tools in the development of their tasks). The instrument allows for the traceability of training needs in digital literacy, as well as the formulation of institutional policies on the use of AI that contribute to the prevention of behaviours associated with academic dishonesty and ensure critical reflection by students on the risks and opportunities of AI in their educational process. Full article

The rapid development of intelligent learning guidance systems has created a favorable environment for personalized learning. By accurately predicting students’ future performance, education can be tailored and teaching strategies optimized. However, traditional prediction algorithms seldom account for highly imbalanced datasets in basic education, overlook temporal factors, and lack further interpretability of the prediction results. To address these shortcomings, we propose Temporal Variational Autoencoder-Generative Adversarial Network (TVAE-GAN), a temporal variational autoencoder-generative adversarial network model aimed at providing early warnings for high-risk students in basic education, with in-depth interpretability analysis of the prediction results to suit the unique context of basic education. TVAE-GAN extracts features from real samples and introduces a Long Short-Term Memory (LSTM) network to capture dynamic features in time series, helping the model better understand temporal dependencies in the data, remember the sequential causal information of students’ online learning, and achieve better data generation performance. Using these features, the generative model generates new samples, and the discriminator model evaluates their quality, producing outputs that closely resemble real samples through training. The effectiveness of the TVAE-GAN model is validated on a collected online basic education dataset while also advancing the timing of interventions in predictions. The performance differences between the proposed method and classic resampling methods, as well as their impact in the educational field, are analyzed, highlighting that misclassification increases teacher workload and affects students’ emotions. Key influencing factors are identified using a decision-tree surrogate model, providing teachers with multidimensional references for academic assessment. Full article

Road traffic accidents at intersections pose a persistent challenge in Riyadh, Saudi Arabia, contributing significantly to public health burdens and economic losses. Traditional statistical approaches often fail to capture the complex, non-linear interactions among geometric design, traffic parameters, and accident severity. This study develops a multi-methodological machine learning framework to predict intersection accident severity using the Equivalent Property Damage Only (EPDO) metric. Historical data (2017–2023) from Riyadh Municipality for 150 high-risk intersections were analyzed, incorporating predictors such as service road distance (SRD), U-turn distance (UTD), median width (MW), peak hour volume (PHV), heavy vehicle percentage (HV%), and injury/frequency counts. Six algorithms, i.e., Decision Tree, Random Forest, Gradient Boosting, Support Vector Machine, Linear Regression, and Artificial Neural Network, were compared using a 70/30 train–test split and k-fold cross-validation in this study. The Gradient Boosting model achieved superior performance (R² = 0.89 with MSE = 63.43 and RMSE = 7.96) and was selected for final deployment. SHAP feature importance analysis revealed minor injuries (MIs), serious injuries (SRIs), and fatalities (FAs) as the most important dominant predictors, with geometric factors (UTD, MW) and traffic composition (HV%) providing actionable infrastructure insights. The model ranked intersections and identified the “Jeddah Road with Taif Road” (predicted EPDO = 137.22) as the highest-risk location. Evidence-based recommendations include enforcing the minimum 300 m U-turn buffers with staggering service road exits ≥150 m and restricting heavy vehicles during peak hours. The scalable framework developed in this study supports the data-driven prioritization of safety interventions and aligns with sustainable urban mobility goals and offers transferability to other metropolitan contexts worldwide. Full article

Background/Objectives: Food insecurity affects college students at nearly twice the rate of US households, with documented impacts on student academic performance, physical and mental health, and socialization. While frameworks exist to conceptualize general food insecurity and food insecurity in specific contexts, researchers and practitioners lack resources to guide system-level responses to food insecurity on college and university campuses and assess those responses. In this study, we aimed to develop and validate a simple yet comprehensive framework for assessing food insecurity responses within the context of higher education. Methods: We adapted an eight-phase process for framework development: (1) map selected data sources within the multidisciplinary literature, (2) read and categorize selected sources, (3) identify and name concepts, (4) deconstruct and categorize concepts based on their features, (5) group similar concepts together, (6) synthesize concepts into a framework, (7) validate the framework using expert panel review, and (8) revise as necessary. Results: The developed Campus Food Aid Self-assessment (CFAS) framework consists of six dimensions: Student Services and Supports; Involvement; Advocacy; Awareness and Culture Efforts; Education and Training; and Research, Scholarship, and Creative Works. Expert panelists (n = 7) reviewed the proposed framework and confirmed the clarity, comprehensiveness, and representativeness of the proposed dimensions, conceptual definitions, and operational variables. Conclusions: With a comprehensive yet accessible structure, the CFAS framework supports the development, coordination, and improvement of campus-based strategies to address food insecurity and support positive student outcomes. Full article

Exercise-derived reactive oxygen species (ROS) are required for mitochondrial and hypertrophic adaptations, creating a practical trade-off: antioxidant strategies may support short-term performance and recovery yet blunt training signals when mis-timed or over-dosed. We performed a structured narrative review informed by transparent database searches of MEDLINE, Scopus, and SPORTDiscus (2000–2025), prioritizing human intervention studies and using mechanistic evidence to interpret plausibility. Evidence was mapped by antioxidant class, dose, timing, training modality, and context. Across trials, chronic high-dose vitamins C/E taken close to key sessions are most consistently associated with attenuation of redox-sensitive signaling, whereas food-first polyphenols and selected bioactives (e.g., tart cherry/anthocyanins, pomegranate, and curcumin) more often support recovery when positioned away from adaptation-critical workouts, without clear evidence of impaired training gains. N-acetylcysteine can acutely improve tolerance to repeated high-intensity exercise, but effects during prolonged training remain uncertain and appear context-dependent. We propose Redox-Adaptive Periodization, aligning antioxidant class, dose, and timing with the primary objective (adaptation vs. immediate readiness) and environmental constraints, and we outline methodological priorities to advance precision redox management. Full article

The precise segmentation of small-sized defects on wood surfaces is critical for the quality grading of glued laminated timber (GLT). Existing semantic segmentation models face core bottlenecks in this context: high miss rates, blurred boundary localization, and excessive size measurement errors. To address these issues, this paper proposes an improved Defect-Mask2Former model that integrates an Attention-Guided Pyramid Enhancement (AGPE) module and a Defect Boundary Calibration and Correction (DBCC) module. Through synergistic optimization, the model achieved pixel-level precise segmentation. To support model training and validation, a custom image acquisition device was designed, and the PlankDefSeg dataset was constructed, comprising 3500 pixel-level annotated images covering five defect types across six industrial wood species. Experimental results demonstrate that on the PlankDefSeg dataset, Defect-Mask2Former achieved a mean Intersection over Union (mIoU) of 85.34% for small-sized defects, a 17.84% improvement over the baseline Mask2Former. The miss rate was reduced from 20.78% to 5.83%, and the size measurement error was only 2.86%, strictly meeting the ≤3% accuracy requirement of the GB/T26899-2022 standard. The model achieved an inference speed of 27.6 FPS, satisfying real-time detection needs. By integrating the model into the GLT grading workflow, a grading accuracy of 94.3% was achieved, and the processing time per timber was reduced from 30 s to 1.5 s, a 20-fold efficiency improvement. This study provides reliable technical support for intelligent GLT quality grading and offers a reference solution for other industrial surface defect segmentation tasks. Full article

This article investigates whether increasing economic status through adopting digital payment capabilities in Delhi fosters economic and financial inclusion among urban women street vendors in Mahila Haat. Digital freedom is a new step forward in technology for everyone. Still, a woman not only balances the social responsibilities of childbearing, caring for her children and family, and struggling with economic issues, health issues, and undernourishment, but can also balance the household job of street vending to increase self-esteem and financial independence. This research work conducted a sampling survey and applied the Kruskal–Wallis H-test with a p-value (0.05) significance level by evaluating 11 variables to investigate the relationship between the digital capabilities and economic independence of street vendors in Mahila Haat (a women’s market where the vendors are all women) in the Red Fort area of New Delhi. UPI systems were created using measurements based on a five-point Likert scale to analyze different levels of satisfaction in clusters of digital capabilities on digital platforms. Further, the ordinary least squares (OLS) method was used to estimate quality of life and social happiness in the context of digital empowerment. Digital payment systems positively influence women’s empowerment. Women vendors can adopt digital payment methods, making them economically independent. The positive relationship between women vendors and customer satisfaction before UPI use and after UPI use is also analyzed. This research will be helpful for both government and non-government organizations to provide financial assistance, informational awareness, skill development training, and advocacy for gender equality to increase women’s empowerment. Full article

Serious video games are digital games designed for purposes beyond entertainment, typically to support education, training, health interventions, or behaviour change. They combine game mechanics with psychological and pedagogical principles, such as feedback, repetition, goal-setting, and scaffolding, to create interactive environments that facilitate learning, skill development, and sustained engagement. In many cases, they are built to simulate realistic tasks or decision contexts, allowing users to practise skills, test strategies, and learn from consequences in a low-risk setting. Within cyberpsychology, serious video games are particularly valuable because they provide structured digital contexts for examining how technology shapes cognition, emotion, motivation, and behaviour. They enable researchers and practitioners to observe how users respond to digital rewards, challenges, social features, and immersive environments, as well as how these features influence outcomes such as self-efficacy, persistence, attention, and emotion regulation. As a result, serious video games operate at the intersection of psychological theory, human–technology interaction, and applied digital intervention design. This entry provides an overview of their development, theoretical foundations, applications, effectiveness, and associated ethical considerations. Full article

This study examines how university type (public vs. private) and disciplinary background influence the adoption of Information and Communication Technologies (ICT) and self-perceived digital competence among university professors in Latin America. Identifying institutional and disciplinary disparities is essential in the context of accelerated digital transformation in higher education. The sample comprised 1114 professors from public and private universities, and data was collected using a validated instrument measuring ICT valuation, frequency of use, and perceived digital competence. Multivariate analyses were conducted to assess differences by institutional type and disciplinary field. The results show significant differences in ICT valuation, usage frequency, and perceived digital competence across university types and disciplines. Professors from private universities reported higher digital preparedness, while disciplinary areas displayed distinct ICT adoption patterns. Although ICT use increased across all groups during the pandemic, the digital gap between public and private institutions narrowed but was not fully eliminated. These findings support the development of targeted professional training, strategic resource allocation, and institutional policies, particularly in public universities, to enhance digital competence and promote sustainable ICT integration, contributing to educational equity and progress toward Sustainable Development Goals. Full article

The modern information retrieval field increasingly relies on hybrid search systems combining sparse retrieval with dense neural models. However, most existing hybrid frameworks employ static mixing coefficients and independent component training, failing to account for the specific needs of individual queries and corpus heterogeneity. In this paper, we introduce an adaptive hybrid retrieval framework featuring query-driven alpha prediction that dynamically calibrates the mixing weights based on query latent representations instantiated in a lightweight low-latency configuration and a full-capacity encoder-scale predictor, enabling flexible trade-offs between computational efficiency and retrieval accuracy without relying on resource-inefficient LLM-based online evaluation. Furthermore, we propose antagonist negative sampling, a novel training paradigm that optimizes the dense encoder to resolve the systematic failures of the lexical retriever, prioritizing hard negatives where BM25 exhibits high uncertainty. Empirical evaluations on large-scale multilingual benchmarks (MLDR and MIRACL) indicate that our approach demonstrates superior average performance compared to state-of-the-art models such as BGE-M3 and mGTE, achieving an nDCG@10 of 74.3 on long-document retrieval. Notably, our framework recovers up to 92.5% of the theoretical oracle performance and yields significant improvements in nDCG@10 across 16 languages, particularly in challenging long-context scenarios. Full article

Predicting loan defaults is a significant challenge for financial institutions; however, current machine learning techniques often encounter issues in areas such as data privacy, cross-institutional cooperation, and model transparency. The restrictions on the practical implementation of advanced predictive models are centralized training paradigms, which limit the application of advanced models because of regulatory and confidentiality issues, and black-box decision making, which diminishes confidence in automated credit risk tools. This study mitigates these problems by adopting a federated-inspired decentralized ensemble learning model combined with explainable artificial intelligence (XAI) in predicting loan defaults. Various machine learning classifiers are trained on partitioned institutional data without the need to share any data; they include K-Nearest Neighbors, support vector machine, random forest, and XGBoost. They use a prediction-level aggregation strategy to simulate the collaborative decision-making process without losing locality of data. SHAP and LIME are used to promote model interpretability by giving both global and local explanations of the consequences of prediction. The proposed framework was tested on a large public dataset of loans that contains more than 116,000 records, including various financial and borrower-related features. The experimental findings show that XGBoost has high and reliable predictive accuracy in both centralized and decentralized scenarios, achieving 99.7% accuracy under federated-inspired evaluation. The explanation analysis shows interest rate spread and upfront charges as the most significant predictors of loan default risk. The main contributions of this research are as follows: (i) a privacy-preserving decentralized ensemble learning framework that is applicable in multi-institutional financial contexts, (ii) a detailed analysis of centralized and decentralized predictive performances, and (iii) the pipeline of the XAI, which can be used to increase its transparency and regulatory confidence in automated credit risk evaluation. These results prove that decentralized learning combined with explainable AI can provide high-performing, transparent and privacy-sensitive loan default prediction systems in practice in real-world banking systems. Full article

Background/Objectives: The increasing volume of laboratory testing and the tightening of quality standards have rendered automation tasks in medical laboratories highly relevant. Conventional total laboratory automation (TLA) systems demonstrate high throughput; however, their economic and organizational efficiency is often constrained by their complex integration and substantial implementation costs. In this context, collaborative robots (cobots) are attracting increasing attention due to their ability to perform pre-analytical and logistical tasks in close association with laboratory personnel. The objective of the present study was the systematic integration of commercially available cobots into the pre-analytical workflow of a large centralized laboratory. Methods: The implemented system incorporated a set of specialized modules, including decapping, barcode orientation and verification, analyzer loading, aliquoting, and specimen sorting, with bidirectional integration into the Laboratory Information System (LIS). The architectural design, control algorithms, and primary effects on labor input and operational turnaround time were evaluated. Results: The results demonstrated that the implementation of cobots into laboratory processes led to an 87% reduction in labor input, a 3.4% improvement in liquid aliquoting accuracy, and an overall improvement in nominal throughput, while requiring minimal personnel training. However, human operators performed the aliquoting procedure significantly faster than cobots, with an average speed advantage of 42.5%. Conclusions: The use of collaborative robotic systems in centralized medical laboratories appears promising due to their operational efficiency and flexibility compared to conventional automation platforms and manual workflows. The effect of the use of cobots on the quality/accuracy of the tests needs to be evaluated, and perhaps a larger study of multiple laboratories needs to be conducted to confirm the results are generalizable. Full article

Accurate water body segmentation from multispectral remote sensing imagery is critical for hydrological monitoring and environmental management. However, leveraging transfer learning with pre-trained models remains challenging due to the dimensional mismatch between three-channel RGB-based architectures and multi-band spectral data. To address this, this study proposes a novel segmentation network, termed Intelligent Gated Fusion Network (IGF-Net), built upon a dual-branch feature encoder module and a core Intelligent Gated Fusion Module (IGFM). The IGFM achieves adaptive fusion of visual and spectral features through a cascaded mechanism integrating differences-and-commonalities parallel modeling, channel-context priors, and adaptive temperature control. We evaluate IGF-Net on the newly constructed Tiangong-2 remote sensing image water body semantic segmentation dataset, which comprises 3776 meticulously annotated multispectral image patches. Comprehensive experiments demonstrate that IGF-Net achieves strong and consistent performance on this dataset, with an Intersection over Union of 0.8742 and a Dice coefficient of 0.9239, consistently outperforming the evaluated baseline methods, such as FCN, U-Net, and DeepLabv3+. It also exhibits strong cross-dataset generalization capabilities on an independent Sentinel-2 water segmentation dataset. Ablation studies and visualization analyses confirm that the proposed fusion strategy significantly enhances segmentation accuracy and stability, particularly in complex scenarios. placeholder Full article

Hyperspectral image (HSI) classification regarding vegetation is hampered by strong intra-class spectral variability and inter-class similarity, and commonly used random pixel splits can introduce spatial-context leakage that inflates test accuracy in patch-based models. To address these issues, we propose a classification framework that couples a leakage-free block partition (LFBP) strategy with class-aware multi-prototype contrastive loss (CAMP-CL). LFBP assigns non-overlapping spatial blocks to training/validation/test sets and reserves a buffer matched to the patch radius to prevent contextual overlap while keeping class distributions balanced. CAMP-CL represents each class with multiple learnable prototypes and performs supervised contrastive learning at the prototype level, encouraging compact yet multimodal intra-class embedding and improved inter-class separation. Experiments conducted on the Matiwan Village airborne HSI dataset under the LFBP protocol show that the proposed method can achieve 91.51% overall accuracy (OA) and 91.49% average accuracy (AA). Compared with the strongest baseline, supervised contrastive learning (SupCon), the proposed method yields consistent gains of 1.07 percentage points (pp) in both OA and AA while improving OA by 5.76 pp over the cross-entropy baseline. The results suggest that CAMP-CL is beneficial for addressing the challenges of HSI classification for fine-grained vegetation, while leakage-free evaluation protocols are important for obtaining more reliable performance estimates in practical settings. Full article

Building Information Modeling (BIM) has changed the landscape of the architectural, engineering, and construction (AEC) industry in recent decades. However, BIM is not well researched in most developing countries; in particular, few studies have addressed its adoption in Malawi. A non-probability, purposive sampling approach was adopted. A total of 143 questionnaires were completed. This research reveals that, while construction experts are aware of BIM, the level of uptake remains quite low. Architects in Malawi are the most knowledgeable, followed by land surveyors and then engineers. This research shows that most experts in Malawi are at level 1 of BIM usage, which is the first stage of BIM adoption and is characterized by the use of 3D models and output representation. Furthermore, the study results have shown that the Malawian AEC sector is currently succeeding at the modelling stage of maturity but is stalled by lack of collaborative frameworks, such as Integrated Project Delivery (IPD). Therefore, unless the industry shifts toward a unified Common Data Environment (CDE), advanced capabilities like clash detection will remain underutilized and disconnected from broader project success metrics. Statistical analysis has shown that the correlation analysis demonstrates a strong link (r = 0.75) between Integrated Project Delivery (IPD) and high BIM maturity, whereas traditional Design-Bid-Build methods show a critical misalignment with digital workflows. The study identifies high software costs and a lack of national standards as the primary barriers to adoption. Therefore, there is a need for robust sensitization to the benefits of BIM and training to improve its uptake in the context of Malawi’s construction industry. In order to advance Malawi’s BIM maturity, the research recommends a strategic shift toward integrated procurement models, the establishment of national BIM mandates, and the modernization of technical education to bridge the existing knowledge gap. Full article