Search Results (272)

Undergraduate dropout is a multidimensional phenomenon with implications for higher education, economic development, and social and cultural transformation, posing complex challenges for society as a whole. To address this, universities require effective dropout risk assessments for both domestic and international students, enabling the implementation of tailored strategies and support. This study sourced a dataset from multiple faculties, comprising 3544 records for domestic students (Portuguese) and 86 for international students, considering 23 features. To balance the data, Conditional Tabular Generative Adversarial Networks were utilized to generate 487 synthetic samples with comparable statistical characteristics for training (85%) while retaining the original 86 real samples for testing (15%), thus maintaining an identical train–test split for evaluating domestic students. An Automated Machine Learning framework, employing ensemble learning algorithms, achieved outstanding performance, with the Light Gradient Boosting Machine proving the most effective for domestic students and Categorical Boosting for international students, both achieving test accuracies exceeding 0.90. The analysis revealed that improving academic performance during the first and second semesters was key to reducing dropout risks. Once a satisfactory level was reached, further improvements had minimal impact. Therefore, the focus should be on achieving satisfactory grades. Other objective identity factors, such as age and gender, were less influential than academic performance. A web-based application incorporating the developed models was created, offering an open-access tool for forecasting dropout risks, with all code made publicly available for further research into undergraduate performance, which could be extended to other nations. Full article

Off-system bridges are critical components of the United States’ transportation infrastructure, providing essential access to rural communities and enabling residents to reach vital services such as employment, education, and healthcare. Many of these bridges are structurally deficient, functionally obsolete, and unmaintained. This disproportionately hinders the mobility of underserved populations, worsening socioeconomic disparities. Despite existing research, there is insufficient focus on the unique challenges posed by off-system bridges, including handling the class imbalanced nature of the bridge condition rating dataset. This paper predicts bridge deck conditions by using Generative Adversarial Networks with Focal Loss (GAN-FL) to generate synthetic data which enhances precision–recall balance in imbalanced datasets. Results show that integrating GAN-FL with random forest (RF) classifiers significantly enhances the performance of minority classes, improving their precision, recall, and F1 scores. The study finds that augmenting training data using GAN-FL greatly enhances minority class prediction, thereby assisting in accurate bridge condition modeling. Full article

This study presents SAIL-Y (Sailing Artificial Intelligence for Learning in Youth), a novel gender-focused recommender system designed to promote female participation in STEM careers through data-driven guidance. Drawing inspiration from the metaphor of an academic journey as a voyage, SAIL-Y functions as a digital compass—leveraging socioeconomic profiles and standardised test results (Saber 11, Colombia) to help students navigate career decisions in high-impact academic fields. SAIL-Y integrates multiple machine learning strategies, including collaborative filtering, bootstrapped data augmentation to rebalance gender representation, and socioeconomic-aware conditioning, to generate personalised and bias-controlled career recommendations. The system is explicitly designed to skew recommendations toward STEM disciplines for female students, countering systemic underrepresentation in these fields. Using a dataset of 332,933 Colombian students (2010–2021), we evaluate the performance of different recommendation architectures under the SAIL-Y framework. The results show that a gender-oriented recommender design increases the proportion of STEM career recommendations for female students by up to 25% compared to reference models. Beyond technical contributions, this work proposes an ethically aligned paradigm for educational recommender systems—one that empowers rather than merely predicts. SAIL-Y is thus envisioned as both a methodological tool and a socio-educational intervention, supporting more equitable academic journeys for future generations. Full article

This paper presents TVAE-SSL, a novel semi-supervised learning (SSL) paradigm that involves Tabular Variational Autoencoder (TVAE)-sampled synthetic data injection into the training process to enhance model performance under low-label data conditions in Educational Data Mining tasks. The algorithm begins with training a TVAE on the given labeled data to generate imitative synthetic samples of the underlying data distribution. These synthesized samples are treated as additional unlabeled data and combined with the original unlabeled ones in order to form an augmented training pool. A standard SSL algorithm (e.g., Self-Training) is trained using a base classifier (e.g., Random Forest) on the combined dataset. By expanding the pool of unlabeled samples with realistic synthetic data, TVAE-SSL improves training sample quantity and diversity without introducing label noise. Large-scale experiments on a variety of datasets demonstrate that TVAE-SSL can outperform baseline supervised models in the full labeled dataset in terms of accuracy, F1-score and fairness metrics. Our results demonstrate the capacity of generative augmentation to enhance the effectiveness of semi-supervised learning for tabular data. Full article

Lately, there has been a notable surge in the use of AI-driven dialogue systems like ChatGPT-3.5 within the realm of education. Understanding the factors that are associated with student engagement in these digital platforms is crucial for maximizing their potential and long-term efficacy. This study aims to systematically identify the key drivers behind university students’ loyalty to ChatGPT. Data gathered from university participants was analyzed using structural equation modeling. The findings indicate that novelty value is positively associated with both task attraction and hedonic value. Perceived intelligence shows significant associations with knowledge acquisition, task attraction, and hedonic value. Moreover, knowledge acquisition is positively related to task attraction and hedonic value, while creepiness is negatively related to them. Both task attraction and hedonic value demonstrate significant relationships with satisfaction and loyalty, with trust also positively associated with satisfaction. These insights provide a clearer understanding of what motivates university students to engage with AI conversational platforms like ChatGPT. This information is invaluable for stakeholders aiming to augment the adoption and effective use of such tools in educational contexts. Full article

In recent years, Location-Based Augmented Reality (LAR) systems have been increasingly implemented in various applications for tourism, navigation, education, and entertainment. Unfortunately, the LAR content creation using conventional desktop-based authoring tools has become a bottleneck, as it requires time-consuming and skilled work. Previously, we proposed an in-situ mobile authoring tool as an efficient solution to this problem by offering direct authoring interactions in real-world environments using a smartphone. Currently, the evaluation through the comparison between the proposal and conventional ones is not sufficient to show superiority, particularly in terms of interaction, authoring performance, and cognitive workload, where our tool uses 6DoF device movement for spatial input, while desktop ones rely on mouse-pointing. In this paper, we present a comparative study of authoring performances between the tools across three authoring phases: (1) Point of Interest (POI) location acquisition, (2) AR object creation, and (3) AR object registration. For the conventional tool, we adopt Unity and ARCore SDK. As a real-world application, we target the LAR content creation for pedestrian landmark annotation across campus environments at Okayama University, Japan, and Brawijaya University, Indonesia, and identify task-level bottlenecks in both tools. In our experiments, we asked 20 participants aged 22 to 35 with different LAR development experiences to complete equivalent authoring tasks in an outdoor campus environment, creating various LAR contents. We measured task completion time, phase-wise contribution, and cognitive workload using NASA-TLX. The results show that our tool made faster creations with 60% lower cognitive loads, where the desktop tool required higher mental efforts with manual data input and object verifications. Full article

Background: Periodontally accelerated osteogenic orthodontics (PAOO) is a surgical procedure to accelerate orthodontic tooth movement and minimize periodontal complications. This study surveyed U.S. periodontists to assess various aspects of the procedure as regards prevalence, training, and execution. Methods: The authors developed a unique questionnaire, the first national study of this type, housed on the Qualtrics^® survey platform, to analyze trends in PAOO training and use. Unique recruitment emails were sent to 3154 members of the American Academy of Periodontology. 449 U.S. periodontists/3154 surveyed (14.2%) responded to this web-based, anonymized survey. IBM statistical software (SPSS V28) was used for data analysis. Results: Among respondents, PAOO training was received during residency (32.7%) and by continuing education (CE) (50.8%), with higher CE (57.3%) by those who did not receive PAOO residency training (p < 0.001). 38.5% of periodontists perform PAOO, and those most likely to perform PAOO had both PAOO residency training and CE, with 78.5% performing 1–5 cases/year. Most (87.7%) received 1–2 PAOO referrals/year from orthodontists or general dentists. Differences in techniques and materials were the type of bone graft or membrane used, the position of corticotomies, and the timing of orthodontic movement. The primary PAOO goal was “rapid tooth movement” (41.1%) and to “increase the alveolar housing” (37.2%). The secondary (38%) and tertiary (37.2%) ranked goals were “augment dehiscence or fenestration”, with the “prevention of apical root resorption” ranked as their quaternary goal. Conclusions: The results of this survey provide data on the trends, training, and use of PAOO among U.S. periodontists. This information may aid in developing residency curriculum and performing PAOO research. Full article

This study analyzes the impact of research and development (R&D) investment on economic growth in Panama, an emerging economy with structural challenges in its innovation system. Using a multivariate econometric approach that included elastic net regularization and fixed-effect panel data estimation, the analysis incorporated key explanatory variables such as public education expenditure, inflation, infrastructure investment, population growth, and exports. The results indicated that both R&D and education spending have a positive and statistically significant effect on GDP growth, while inflation has a negative impact and exports show no significant effect. To ensure robustness, the study applied the augmented Dickey–Fuller test for stationarity, nonparametric bootstrapping (1000 replications), and multiple diagnostic tests, including RMSE, adjusted R², Durbin–Watson statistic, and White’s test. Scenario-based projections suggest that gradual and sustained increases in R&D investment, supported by stronger institutional coordination and absorptive capacity, could enhance Panama’s long-term productivity and innovation outcomes. The findings underscore that improving R&D funding alone is not sufficient; effective governance and coherent science, technology, and innovation (STI) policies are essential. This research contributes empirical evidence to a relatively underexplored area in the development literature and offers strategic insights for policymakers seeking to build more integrated and sustainable STI ecosystems in emerging economies. Full article

Background: Large Language Models (LLMs) have driven significant advances in artificial intelligence (AI), with transformative applications across numerous scientific fields, including biomedical research and drug development. However, despite growing interest in adjacent domains, their adoption in pharmacometrics, a discipline central to model-informed drug development (MIDD), remains limited. This study aims to systematically explore the potential role of LLMs across the pharmacometrics workflow, from data processing to model development and reporting. Methods: We conducted a comprehensive literature review to identify documented applications of LLMs in pharmacometrics. We also analyzed relevant use cases from related scientific domains and structured these insights into a conceptual framework outlining potential pharmacometrics tasks that could benefit from LLMs. Results: Our analysis revealed that studies reporting LLMs in pharmacometrics are few and mainly limited to code generation in general-purpose programming languages. Nonetheless, broader applications are theoretically plausible and technically feasible, including information retrieval and synthesis, data collection and formatting, model coding, PK/PD model development, support to PBPK and QSP modeling, report writing and pharmacometrics education. We also discussed visionary applications such as LLM-enabled predictive modeling and digital twins. However, challenges such as hallucinations, lack of reproducibility, and the underrepresentation of pharmacometrics data in training corpora limit the actual applicability. Conclusions: LLMs are unlikely to replace mechanistic pharmacometrics models but hold great potential as assistive tools. Realizing this potential will require domain-specific fine-tuning, retrieval-augmented strategies, and rigorous validation. A hybrid future, integrating human expertise, traditional modeling, and AI, could define the next frontier for innovation in MIDD. Full article

Assessing teaching behavior is essential for improving instructional quality, particularly in Physical Education, where classroom interactions are fast-paced and complex. Traditional evaluation methods such as questionnaires, expert observations, and manual discourse analysis are often limited by subjectivity, high labor costs, and poor scalability. These challenges underscore the need for automated, objective tools to support pedagogical assessment. This study explores and compares the use of Transformer-based language models for the automatic classification of teaching behaviors from real classroom transcriptions. A dataset of over 1300 utterances was compiled and annotated according to the teaching styles proposed in the circumplex approach (Autonomy Support, Structure, Control, and Chaos), along with an additional category for messages in which no style could be identified (Unidentified Style). To address class imbalance and enhance linguistic variability, data augmentation techniques were applied. Eight pretrained BERT-based Transformer architectures were evaluated, including several pretraining strategies and architectural structures. BETO achieved the highest performance, with an accuracy of 0.78, a macro-averaged F1-score of 0.72, and a weighted F1-score of 0.77. It showed strength in identifying challenging utterances labeled as Chaos and Autonomy Support. Furthermore, other BERT-based models purely trained with a Spanish text corpus like DistilBERT also present competitive performance, achieving accuracy metrics over 0.73 and and F1-score of 0.68. These results demonstrate the potential of leveraging Transformer-based models for objective and scalable teacher behavior classification. The findings support the feasibility of leveraging pretrained language models to develop scalable, AI-driven systems for classroom behavior classification and pedagogical feedback. Full article

This Scoping Review (ScR) synthesizes advances in architectural eye-tracking (ET) research published between 2010 and 2024. Drawing on 75 peer-reviewed studies that met clear inclusion criteria, it monitors the field’s rapid expansion, from only 20 experiments before 2018, to more than 45 new investigations in the three years thereafter, situating these developments within the longer historical evolution of ET hardware and analytical paradigms. The review maps 13 recurrent areas of application, focusing on design evaluation, wayfinding and spatial navigation, end-user experience, and architectural education. Across these domains, ET reliably reveals where occupants focus, for how long, and in what sequence, providing objective evidence that complements designer intuition and conventional post-occupancy surveys. Experts and novices might display distinct gaze signatures; for example, architects spend longer fixating on contextual and structural cues, whereas lay users dwell on decorative details, highlighting possible pedagogical opportunities. Despite these benefits, persistent challenges include data loss in dynamic or outdoor settings, calibration drift, single-user hardware constraints, and the need to triangulate gaze metrics with cognitive or affective measures. Future research directions emphasize integrating ET with virtual or augmented reality (VR) (AR) to validate design interactively, improving mobile tracking accuracy, and establishing shared datasets to enable replication and meta-analysis. Overall, the study demonstrates that ET is maturing into an indispensable, evidence-based lens for creating more intuitive, legible, and human-centered architecture. Full article

Eye tracking, a fundamental process in gaze analysis, involves measuring the point of gaze or eye motion. It is crucial in numerous applications, including human–computer interaction (HCI), education, health care, and virtual reality. This study delves into eye-tracking concepts, terminology, performance parameters, applications, and techniques, focusing on modern and efficient approaches such as video-oculography (VOG)-based systems, deep learning models for gaze estimation, wearable and cost-effective devices, and integration with virtual/augmented reality and assistive technologies. These contemporary methods, prevalent for over two decades, significantly contribute to developing cutting-edge eye-tracking applications. The findings underscore the significance of diverse eye-tracking techniques in advancing eye-tracking applications. They leverage machine learning to glean insights from existing data, enhance decision-making, and minimize the need for manual calibration during tracking. Furthermore, the study explores and recommends strategies to address limitations/challenges inherent in specific eye-tracking methods and applications. Finally, the study outlines future directions for leveraging eye tracking across various developed applications, highlighting its potential to continue evolving and enriching user experiences. Full article

Background: Augmented reality (AR) is revolutionizing implant and tooth-supported prosthodontics (ITSP) through enhanced precision, workflow efficiency, and educational outcomes. This scoping review systematically evaluates AR’s clinical applications, educational impacts, and implementation challenges. Methods: Following PRISMA-ScR guidelines, comprehensive searches were conducted in PubMed, Scopus, Web of Science, and Embase (2015–2025) for peer-reviewed studies on AR in ITSP. Eighteen studies met inclusion criteria after dual independent screening. Data extraction focused on clinical outcomes, educational benefits, and technological limitations. Results: AR applications demonstrated: ITSP Practice: Submillimeter implant placement accuracy (0.42–0.69 mm entry deviation; p < 0.001 vs. freehand), 30% faster intraoral scanning (44 s vs. 63 s), and 37% reduction in preparation errors (p < 0.05); ITSP Education: 22–30% faster skill acquisition (p < 0.05) and 99% reduction in assessment time (10.5 s vs. 2 h/case). Key Gaps: Limited to two randomized controlled trials (RCTs), hardware costs ($3500–$10,000), and lack of standardized validation protocols. Conclusions: While AR significantly enhances ITSP precision and training efficiency, widespread adoption requires longitudinal clinical validation, cost-effectiveness analyses, and interoperable digital workflows. Full article

The Italian school building stock is largely outdated, with structural and technological inadequacies leading to low comfort and high energy consumption. Addressing this challenge requires large-scale renovation supported by an integrated, data-driven approach. This study conducted a nationwide analysis of over 40,000 school buildings. After incomplete or inconsistent records were filtered out, a refined subset was selected. Building forms were reconstructed by cross-referencing GIS data with multiple open data sources. Using supervised machine learning, the research identifies and classifies recurring morphological patterns to define a set of 3D school building archetypes. These archetypes are enriched with spatial configurations and physical characteristics aligned with national educational standards. The result is a macrotypological classification based on form, conceived as part of an operational tool to support policymakers, designers, and public administrations in selecting effective retrofit strategies. This contributes to the creation of large-scale national renovation strategies, as well as Renovation Roadmaps and Digital Building Logbooks in line with the Energy Performance of Buildings Directive (EPBD IV), specifically tailored to the Italian context. The novelty of this work lies in its unprecedented scale and the use of AI to enable fast, replicable assessments of retrofit potential, thereby supporting informed decisions in energy-efficient renovation planning. Full article

Autonomous ground vehicles (AGVs) are of major importance in exploration missions since they perform difficult tasks in changing or harmful environments. Mapping and exploration is crucial in hazardous areas, or areas inaccessible to humans, demanding autonomous navigation. This paper proposes a lightweight, low-cost AGV platform, which will be used in resource-constrained situations and aimed at scenarios like exploration missions (e.g., cave interiors, biohazard environments, or fire-stricken buildings) where there are serious security threats to humans. The proposed system relies on simple ultrasonic sensors when navigating and applied traversal algorithms (e.g., BFS, DFS, or $A^{*}$) during path planning. Since on-board microcomputers have limited memory, the traversal data and direction decisions are stored in a file located on an SD card, which supports long-term, energy-saving navigation and risk-free backtracking. A fish-eye camera set on a servo motor captures three photos ordered from left to right and stores them on the SD card for further off-line processing, integrating each frame into a low-frame-rate video. Moreover, when the battery level falls below 50%, the exploration path does not extend further and the AGV returns to the base station, thus combining a secure backtracking procedure with energy-efficient decisions. The resultant platform is low-cost, modular, and efficient at augmenting; thus it is suitable for exploring missions with applications in search and rescue, educational robotics, and real-time applications in low-infrastructure environments. Full article