Knowledge-guided graph machine learning for spatially distributed prediction of daily discharge and nitrogen export dynamics.

Exploring the mechanism of analogy and explicit instructions on self-efficacy, performance, and learning of golf putting task: Analysis of mental representation.

Using spatially explicit machine learning to enhance assessment of the Global Gravity-based Groundwater Product for groundwater storage change in Germany

An explicit physics-residual learning model for vehicle longitudinal trajectory prediction

GeoShapley-based interpretation of older adult pedestrian fatal vs injury crash frequency in dense urban environments.

The increasing release of greenhouse gases (GHGs) like CO₂, CH₄, N₂O, and industrial contaminants (indirect GHGs) such as SO₂ and H₂S has prompted significant global worries due to their role in climate change, air pollution, and harm to the environment. Ionic liquids (ILs) as green solvents have emerged as promising alternatives to traditional solvents because of their minimal volatility, high thermal stability, and adjustable physicochemical characteristics. Yet, limited gas solubility data in ILs is hindering their applications in carbon capture and air pollution control. Machine learning (ML) is a powerful tool for modeling and simulating the solubility of polluting gases in ILs. This research aims to critically review recent progress in ML modeling of pollutant gas removal by ILs. More importantly, a new ML model of genetic programming (GP) was developed to generate an explicit and accurate mathematical equation to predict the solubility of SO 2 , CH 4 , N 2 O, CO, H 2 S and CO 2 in ILs, using a large dataset (3209) for different gas-IL systems. Using temperature, pressure, and structural related parameters of ILs and gases as input parameters, the model achieved a high accuracy (R 2 > 0.97). Finally, a simple Excel method for calculating gas solubility has been created for prediction and modeling purposes.

MolUNet++: Adaptive-grained explicit substructure and interaction aware molecular representation learning

Sensorimotor adaptation depends on implicit recalibration and explicit strategy. These processes are commonly assumed to sum (A = I + E), and this additivity assumption justifies subtractive measurement and informs computational models of motor learning. Recent work has challenged additivity by examining regression slopes between implicit and explicit measures. When slopes deviate from β = -1, the interpretation has been that the processes are "sub-additive" and fail to sum as expected. Here, we show this reasoning is mistaken. Regression slopes reflect covariance structure: how learning processes relate across individuals. Additivity is a claim about motor output combination: whether learning processes sum within individuals. These are different questions, and regression slopes do not address the latter. We derive the expected slope under subtractive logic and show it equals β = -1 only when total adaptation is uncorrelated with the measured component. Monte Carlo simulations confirm this benchmark is routinely rejected under realistic covariance structures, even when additivity is enforced. Under independent measurement of the learning processes, the regression slope depends on covariance structure, in which additivity does not constrain. Thus, there is no regression slope benchmark for diagnosing additivity. Moreover, the regression slopes reported in previous studies fall within the range predicted by shared-error models that adhere to the additivity assumption. Regression slopes do not test additivity; they only indicate how implicit and explicit learning covary across individuals. Challenging the additivity assumption will require direct tests of motor output combination and formal model comparison.

TWT: Textual white-box transformer for natural language understanding.

Implicit (unconscious) learning ostensibly affects cognitive and social skills in both typical and atypical populations, such as those on the autism spectrum. Research into implicit learning in autism has yielded conflicting results, underscoring the need to explore factors that might influence their implicit learning. One such factor is processing style, specifically processing biases for either global (holistic) or local (detail-oriented) processing. In our experiment, we investigated the potential role of processing differences in implicit (and explicit) learning performance in individuals with autism (n = 20) and typically developing (TD) individuals (n = 22), by using a global-local version of the artificial grammar learning (AGL) task. Overall AGL performance and explicit knowledge yielded only a trend toward an interaction suggesting a greater global processing advantage in TD participants compared with that in participants with autism but no conclusive evidence. The above interaction was further observed in terms of implicit knowledge, with TD participants demonstrating higher levels of implicit structural knowledge than individuals with autism during global processing. Implicit knowledge between group differences during local processing remains weak/inconclusive. Overall, our findings suggest interesting potential processing differences in the implicit learning between individuals with and without autism.

This paper presents an empirical-experimental study examining whether, and to what extent, undergraduate students of Political Science at an Italian university acquire English socio-political lexis through informal exposure to online media. The study was conducted at the University of Pisa and involved a small sample of 31 second-year students, tested prior to their attendance of the official English Language course, in order to assess their extramural knowledge of specialised and semi-specialised vocabulary related to socio-political themes. Grounded in a broad conception of informal language learning as incidental, self-directed contact with the L2 in non-institutional contexts, the research hypothesises that regular engagement with authentic English-language media may contribute to the development of domain-specific lexical knowledge, even in the absence of an explicit learning intention. Data collection instruments included a placement test, a questionnaire on students’ online media habits, a general vocabulary test, and a purpose-built specialised vocabulary test. Preliminary results indicate that performance on the specialised vocabulary test is more strongly associated with learners’ overall English proficiency level than with the reported quantity of online media exposure. These findings offer initial insights into the role of informal media exposure in the acquisition of disciplinary lexis at university level.

Spatiotemporal prediction (STP) is a fundamental research topic in GIScience, providing estimations of unobserved phenomena across space and time. Geospatial effects—spatial dependence, heterogeneity, and geographical similarity—underpin established spatially explicit learning frameworks, namely spatial dependence learning, regional learning, and transfer learning. A theoretical framework is lacking, however, to quantify these effects’ contributions to the learning process and to guide their optimal utilization in a spatially explicit learning framework construction. To fill this gap, we develop a novel predictability framework that integrates STP learning with geospatial effects. In this framework, we propose geographic entropy methods to quantify data predictability under geospatial effects. Using these predictability values, we construct predictability structures (autocorrelation, stratification, and teleconnection). Then, a knowledge graph and its corresponding knowledge services are established from these predictability structures. The knowledge services provide predictability knowledge to optimize spatially explicit learning frameworks, thereby enhancing STP performance. Experiments on a human-activity data set show that our framework significantly improves spatially explicit learning frameworks, achieving average accuracy gains of 4.40 percent for dependence learning, 20.99 percent for regional learning, and 18.21 percent for transfer learning.

Universal Design for Learning (UDL) offers a framework for fostering inclusive education through multiple means of engagement, representation, and action and expression. This qualitative case study explores how UDL strategies influence participation, engagement, and learning outcomes in a professional undergraduate course redesigned to align with UDL principles. The course included explicit learning objectives, flexible assessments, and multimodal instructional materials. Seven full-time undergraduate students in their final year participated, most identifying as women, with three disclosing a physical or developmental disability and prior experience with academic accommodations. While some participants were familiar with UDL, most had not encountered its principles before the course. Data collection included demographic questionnaires, semi-structured interviews, and classroom observations using the UDL Observation Measurement Tool. Findings suggest that UDL strategies promoted interactive learning, accessibility, and student autonomy. Participants emphasized the benefits of recorded lectures, structured discussions, and flexible assessments in reducing barriers. Several participants noted that UDL fostered a more inclusive learning environment, allowing them to engage without requiring formal accommodations. This study highlights the positive impact of UDL-informed design in professional education, demonstrating how inclusive strategies can enhance engagement and retention. The findings reinforce the value of embedding UDL into curriculum development to support equitable, student-centred learning experiences.

DrowsyDG-Phys: Generalizable driver drowsiness estimation in conditional automated vehicles using physiological signals.

Abstract To the extent that individual financial literacy is important, the expected declining share of students attending and completing college suggests high schools will have an increasingly outside role in imparting individual financial literacy. In this paper, we consider whether high school only financial education enhances financial literacy and knowledge, and whether it has a beneficial impact on adult income, securities ownership, and home ownership. With data from the National Financial Capability Study, we estimate high school financial education only treatment effect parameters of outcome specifications that mitigate the effect of unobservables and selection into treatment. Our parameter estimates reveal that high school financial education has favorable treatment effects—possibly causal—on financial literacy and self‐reported financial knowledge. With respect to income and individual decisions to invest in securities and purchase a home, high school only financial education has no overall treatment effect. Our results suggest that high school financial education may not sufficiently impart that today's youth will live longer than previous cohorts, with a longer period of post‐retirement life that requires more consumption. This may fail to induce participants in high school financial education to set goals for investing in assets and careers that can finance post‐retirement consumption. Understanding financial strategies for financing post‐retirement consumption should perhaps be made an explicit learning outcome in a strengthened high school financial education curriculum.

Professionalism in gastronomy is often judged through visible outputs – brigade leadership, menu design, and guest experience – while chefs’ ingredient work remains largely backstage. Existing research has been less explicit about how chefs qualify ingredients as professional judgment under real-time service constraints. This study defines chefs’ knowledge of qualified ingredients and identifies two complementary evaluative logics: naturalistic quality in cookery and built-in quality in pastry and bakery. Drawing on a qualitative field study involving six purposively selected chefs working in hotel and independent operations, we conducted one-to-one, face-to-face interviews in 2024, transcribed them verbatim, and analysed the data thematically through manual coding. Chefs treated ingredient selection not as a routine purchasing task but as qualification work resolved through kitchen practice. Labels, brands, and price informed initial screening, while performance in use determined acceptance – sensory behaviour in context for cookery, and functional performance and repeatability within pastry-and-bakery production systems such as doughs, batters, creams, and fillings. The analysis also shows how these logics converge in hybrid judgement when service tightens margins, clarifying professionalism as the sustained labour of making ingredient decisions defensible so that technical integrity and intended sensory character can both hold under pressure. The study reframes ingredient selection as situated professional judgement and supports the inclusion of ingredient qualification as an explicit learning outcome in vocational culinary education.

The rapid growth of digital health research, involving wearable devices, mobile apps, and sociotechnical health systems, raises complex ethical, legal, and social considerations. While institutional review boards and research ethics frameworks address some concerns, less is known about how students and trainees in digital health are systematically educated to recognize and navigate these challenges. Understanding the scope and content of ethics training is critical to ensuring the responsible development and application of digital health technologies. This study investigated how college students are trained to identify and address ethical considerations in digital health research through an analysis of formal curricula and expert perspectives. Researchers reviewed 132 syllabi from 76 academic programs across 62 universities and conducted semistructured interviews with 6 leading digital health scholars. All syllabi were coded for instructional content and learning objectives. Researchers conducted open coding and collaboratively applied affinity diagramming to organize the data into hierarchical themes. All syllabi included instructional content, and most included explicit learning objectives. Analysis identified 7 key themes, which captured both explicit knowledge imparted through formal instruction and tacit knowledge cultivated through laboratory work, mentorship, and applied experiences. Findings highlighted gaps between formal ethics instruction and the realities of research practice. Ethics education in digital health research develops through the interplay of formal coursework and practice-based training, each fostering complementary skills needed for data-intensive and collaborative environments. Together, these pathways support students in identifying ethical issues, applying principles contextually, anticipating emerging risks, and communicating across disciplines; however, access to experiential learning opportunities remains inconsistent. Strengthening ethics training will require expanding structured early research engagement, cultivating communities of practice, and translating tacit ethical reasoning into accessible resources. Integrating ethical reflection into routine research activities may better prepare future digital health researchers to responsibly design and govern sociotechnical health systems.

This study presents a Geospatial Artificial Intelligence (GeoAI) framework for high-resolution Zoonotic Cutaneous Leishmaniasis (ZCL) risk mapping, correlation analysis, and scenario-based projection, integrating geographic information systems (GIS), remote sensing, and neural network architecture. Historical disease maps and multi-temporal satellite-derived environmental layers were jointly modeled using a multilayer perceptron (MLP), two-dimensional convolutional neural networks (2D-CNNs), and three-dimensional CNNs (3D-CNNs). The principal methodological contribution is the implementation of a 3D-CNN, which enables explicit learning of spatiotemporal transmission dynamics. Environmental-disease relationship analyses, based on Pearson coefficients and regression models, identified temperature as the dominant positive environmental driver of ZCL risk. Model performance assessment using root mean square error (RMSE), mean absolute error (MAE), and the area under the receiver operating characteristic curve (AUC) indicates that the 3D-CNN consistently outperforms alternative architectures in capturing complex spatial and temporal patterns. Elevated risk was concentrated in warmer western and southern regions, whereas cooler northern and eastern mountainous areas exhibited lower susceptibility. By 2030, ZCL risk is projected to undergo a spatial shift, with risk decreasing in western regions and intensifying in southern areas, which has direct implications for targeted surveillance and intervention efforts.

Purpose To explore the similarities and differences in clinical practice and intervention choices of Bobath- and non-Bobath-educated physiotherapists during a single post-stroke clinical session. Methods A qualitative interpretive description approach was used, incorporating stimulated recall with video-recorded treatment sessions and in-depth interviews. Thirty-two neurological physiotherapists from diverse clinical settings participated, including 16 Bobath educated (BG) and 16 non-Bobath educated (NBG) therapists. Video recordings provided primary data for intervention interpretation and were cross-referenced with interview data describing clinical reasoning. The Rehabilitation Treatment Specification System (RTSS), including treatment targets, ingredients, and mechanisms of action, was applied to clarify intervention selection and theoretical underpinnings. Results A total of 2507 interventions were analysed and categorised. The BG emphasised sensory stimulation, manual facilitation, and environmental organisation to influence body configuration, postural control, and sensorimotor integration, reflecting implicit neurophysiological mechanisms. The NBG prioritised task repetition, strength training, and verbal cueing, aligning practice with explicit motor learning principles. The BG used therapeutic touch more frequently (65.4%) and structured session around part-task progression, whereas the NBG employed more hands-off strategies (44.3%) and favoured task repetition. Conclusion The RTSS framework effectively illustrated differences in intervention structure, intent, clinical reasoning, and mechanisms of action between Bobath- and non-Bobath-educated physiotherapists.

ABSTRACTVocabulary knowledge is foundational to educational success, but significant gaps exist between students with reading disabilities or those from disadvantaged backgrounds and their peers. These gaps have cascading effects, as children with lower vocabulary knowledge are less likely to acquire new words through independent reading and are less responsive to vocabulary instruction methods like read‐alouds and explicit teaching. The effectiveness of explicit instruction relies on individualization, which typically places substantial demands on educators and thereby hinders the adoption of evidence‐based methods. A potential solution is using audiobooks supplemented by explicit and individualized remote instruction from paraprofessionals. We conducted a randomized controlled trial (RCT) intervention study in which children listened to text‐supplemented audiobooks, either alone or with scaffolded instructional support. Third and fourth‐grade students (N = 314, age: mean (SD) = 9.47(0.57) years) were randomly assigned to one of three conditions (Audiobooks‐Only, Audiobooks+Scaffold, or active control) for 8 weeks. Participants in the two audiobook intervention groups showed significant improvements in book‐specific vocabulary, while the active control group showed no improvement. The effectiveness of the intervention varied by reading ability and socioeconomic status (SES): poor readers benefited only when audiobooks were paired with one‐on‐one scaffolding, whereas children from lower‐SES backgrounds showed modest, nonsignificant gains from audiobook access alone and did not experience additional benefits from scaffolding. Additionally, the Audiobooks+Scaffold group spent more time listening to recommended audiobooks during the study. These findings suggest that text‐supplemented audiobooks, particularly when combined with personalized support, can be a valuable tool for supporting vocabulary development in struggling readers.SummaryChildren successfully learned new vocabulary words by engaging with text‐supplemented audiobooks.Vocabulary gains were largest amongst students who additionally received one‐on‐one remote scaffolding sessions throughout the intervention period.Poor readers only benefited when text‐supplemented audiobooks were paired with one‐on‐one instructional support.Students from lower socioeconomic backgrounds showed smaller, nonsignificant gains from either component, suggesting a need for additional support to achieve comparable vocabulary growth.