Search Results (1,054)

Background/Objectives: Chronic hepatitis C (CHC) infection is an independent risk factor for developing type 2 diabetes mellitus (T2DM). However, it is unknown if antiviral treatment, especially with direct-acting antivirals (DAAs), changes long-term glycemic outcomes. Methods: We conducted a retrospective comparative cohort study of 2489 patients with chronic hepatitis C (CHC) in southern Taiwan between 2005 and 2022 who underwent treatment with either an interferon (IFN)-based or direct-acting antiviral agent (DAA) regimen. Given the distinct treatment eras of IFN and DAA therapies, potential temporal confounding was considered in the analytical design. Patients with existing diabetes or co-infections were excluded. The incidence of new-onset T2DM and longitudinal HbA1c levels were compared between treatment groups over a mean follow-up period of 2.56 years. Results: DAA-treated patients demonstrated a lower crude cumulative incidence of T2DM compared with IFN-treated patients (2.46% vs. 6.91%). However, adjusted analyses did not demonstrate a statistically significant difference between treatment groups. The cumulative risk appeared to plateau after the third year among DAA recipients. Post-therapy, HbA1c levels remained stable in both groups at between 5.5% and 6.5% over as long as five years. Splitting regression revealed that BMI ≥ 30 kg/m², and not treatment type or achieved SVR, was an independent T2DM risk factor. The lowest rates of diabetes incidence were associated with pan-genotypic DAA regimens. Conclusion: DAA-treated patients showed lower crude T2DM incidence than IFN-treated patients; however, this difference was not consistently significant after adjustment for baseline factors. Viral eradication may be associated with favorable metabolic trends; however, the present findings do not establish a causal protective effect against incident T2DM. While increased BMI remained an independent predictor of post-treatment diabetes risk. Full article

Vision-and-language navigation (VLN) traditionally relies on explicit reasoning chains, which, despite being interpretable, impose severe constraints on inference efficiency and scalability in long-range environments. Existing multimodal large language models (MLLMs) frequently encounter latency bottlenecks due to the generation of verbose textual narratives during decision-making. To address these limitations, we propose spatial reasoning vision-and-language navigation (SR-VLN), a novel framework that shifts the paradigm from explicit chain-of-thought (CoT) to an implicit spatial representation space. SR-VLN introduces a pyramidal hierarchical history framework integrated with perceptual compression to condense historical trajectories into multi-scale representations, effectively minimizing token overhead while preserving critical spatial semantics. Rather than generating verbose textual reasoning steps, SR-VLN employs compact, learnable spatial tokens (S-Tokens) to perform agile inference directly within the latent feature space. To establish robust causal mappings between these implicit states and navigational actions, we employ a hybrid training strategy that combines sparse reward supervision with reinforcement learning via GRPO. Extensive evaluations on the R2R, REVERIE, and SOON datasets demonstrate that SR-VLN achieves state-of-the-art overall navigation performance, while maintaining a comparable balance between accuracy and efficiency. Compared to explicit reasoning baselines, our method reduces token consumption by 68% and achieves a 4.1× speedup in inference while reaching a 76.02% success rate and a 73.80% SPL on the R2R unseen split, thereby facilitating near-real-time action prediction in long-range navigation environments. Full article

Optimization of nitrogen (N) management is critical for enhancing maize (Zea mays L.) productivity while maintaining soil health. The present study investigated the impact of split-application fertilization strategies on soil chemical properties and grain yield across three distinct soil types (calcaric fluvisol, luvic phaeozem, and stagnic phaeozem) in Mureș County, Romania, over three cropping seasons (2022–2024). Three fertilization variants were evaluated: the first treatment, designated V1, involved the application of 300 kg/ha NPK 20-20-0 + 300 kg/ha urea, the second treatment V2 utilized 300 kg/ha NPK 20-20-0 + 300 kg/ha NAC 27 N-calcium ammonium nitrate, and the third treatment V3 served as the baseline control, receiving (300 kg/ha NPK 20-20-0). Results indicated that significant differences were observed among the three experimental sites representing contrasting soil types for soil chemical properties and maize productivity. Calcaric fluvisol exhibited the highest production potential, attaining a mean yield of 11,702.78 kg/ha. The impact of N supplementation on soil N levels and maize yield was found to be significant. The variant receiving urea supplementation (V1) achieved the highest median yield of 9560 kg/ha in comparison to the 7420 kg/ha obtained in the control. A strong positive correlation was observed between N index and yield across all soil types (ρ = 0.93 to 0.97, p < 0.001). Fertilization significantly influenced soil pH, CaCO₃ content, nitrogen index, phosphorus availability, and maize yield, whereas humus content remained relatively stable among treatments. These findings indicate that a split-fertilization regime combining NPK with urea provides a favorable balance between productivity and cost-effectiveness and maize output in the Transylvanian Plateau. Full article

Rail-transit station areas concentrate commercial services, employment, and intensive land development, but their vitality is shaped by multiple built-environment conditions rather than rail accessibility alone. Focusing on 20 stations along the Osaka Metro Midosuji Line in Japan, this study uses Japanese chome units, which are small neighborhood-level address and statistical units, within an 800 m pedestrian catchment as analytical units and measures commercial-service agglomeration intensity, employment intensity, and commercial–employment composite vitality. The composite indicator measures the static co-concentration of commercial-service provision and employment carrying capacity, with pedestrian flow, consumption activity, and dwell time treated as separate dimensions of station-area vitality. Ten station-area environmental variables are examined using ordinary least squares (OLS), Lasso, Random Forest, Back-Propagation (BP) Neural Network, and extreme gradient boosting (XGBoost) models, with Shapley additive explanations (SHAP) applied to interpret variable contributions and nonlinear responses. Results show that nonlinear models generally outperform linear models. Development intensity, officially assessed land price, and network distance to the nearest metro station are the most influential variables, showing threshold, marginal, and non-monotonic effects. Split models indicate that commercial-service agglomeration is more sensitive to rail proximity and street-network conditions, whereas employment intensity is more associated with development intensity and land price. These findings support fine-grained station-area renewal and mixed-function planning. Full article

Grading open-ended student work consistently remains a persistent challenge in higher education, and the recent rise of large language models (LLMs) has renewed interest in rubric-guided automated scoring. However, a key gap remains: most studies report correlation rather than agreement, rarely benchmark models against a local human–human baseline, and seldom test whether simple post hoc calibration improves operational fit. This study addresses that gap by examining whether a rubric-guided LLM can approximate local human grading practice for text-based responses in three university courses, using agreement-oriented rather than correlation-only evidence. A total of 930 student responses from Prompt Engineering, Photoshop Design, and AI Video Production were scored by two human raters and by ChatGPT using the same five-criterion analytic rubric (Accuracy, Logical Flow, Specificity, Quality, and Originality; 0.0–3.0 each; Total 0–15). Human consensus (HC) was defined as the mean of the two human scores and was treated as a pragmatic reference rather than a ground truth. Pairwise agreement among H1, H2, AI, and HC was evaluated using ICC(3,1), Pearson correlations, mean absolute error (MAE), Bland–Altman bias and limits of agreement (LoA); a course-specific held-out calibration analysis was additionally conducted. For the Total score, human–human agreement was strong (ICC = 0.819 [0.797, 0.839]). AI–H1 and AI–H2 Total-score agreement were ICC = 0.700 [0.666, 0.732] and 0.767 [0.739, 0.792], respectively, while AI–HC agreement was ICC = 0.763 [0.735, 0.789], with MAE = 1.603 and LoA = [−4.246, 4.045]. At the trait level, AI–HC ICCs exceeded H1–H2 ICCs for all five rubric dimensions, although Quality remained weakly defined in the human baseline. On a 70/30 held-out test split, a course-specific linear calibration modestly improved Total-score ICC from 0.774 to 0.782 and reduced MAE from 1.624 to 1.215, narrowing the LoA from [−4.290, 4.188] to [−3.157, 3.329]. However, threshold-adjacent agreement remained imperfect after calibration. The principal contribution is a conservative, multi-metric agreement benchmark of rubric-guided LLM scoring against a local human baseline, together with a held-out calibration test that informs deployment. The findings concern written responses only and support a conservative conclusion: rubric-guided LLM scoring can assist human grading under fixed local rubrics, but the current evidence supports calibrated human–AI co-grading rather than unsupervised replacement. Full article

Machine-learned molecular force fields require many-body geometry, but obtaining it through Clebsch–Gordan tensor products is computationally expensive. For a strong no-Clebsch–Gordan backbone such as GotenNet, we ask whether the limitation in handling three-body geometry is one of representational capacity or one of training supervision, and separate the two factors with three controlled probes on a single-seed, paper-aligned rMD17 aspirin split. (i) While frame projection of tensor features is comparable to scalar cos-angle triplet cross-attention (SCTA) at pilot scale, algebraically its diagonal scalar collapses to a frame-independent inner product and the remaining channel is parity-odd, making SCTA’s cos-angle input the principled $\mathrm{O}\left(3\right)$ scalar choice. (ii) SCTA matches GotenNet’s converged force accuracy within ∼0.4% without independent gain, indicating that three-body representational capacity is not the binding constraint. (iii) A graph-level auxiliary loss on bond-angle and dihedral statistics gives the best force mean absolute error (MAE; $0.1280$ vs. $0.1303$ kcal/mol/Å) and reduces epochs-to-validation-target by 26–55%. Cross-molecule probes do not extend this finding; a paired salicylic acid comparison shows a directional degradation that, under a configuration-level paired block bootstrap, is significant and opposite in sign to the aspirin effect. Across three random seeds, the auxiliary force-MAE gain is small and seed-dependent but consistently reduces seed-to-seed variance and accelerates convergence, indicating that low-cost three-body supervision can be a more effective lever than added three-body capacity. Full article

The rapid evolution of online learning environments has generated diverse and complex data ecosystems. Recommender systems play a central role in leveraging such heterogeneous data to support personalised learning experiences. However, many deep learning-based recommender systems still rely on identifier-based representations that capture co-occurrence and collaborative patterns while overlooking the semantic information embedded in educational activities and the temporal dynamics of learner behaviour. To address these limitations, this study proposes a collaborative-enhanced semantic–sequential recommendation framework for educational platforms that combines structured semantic representation learning, sequential behavioural modelling, and collaborative preference modelling. The proposed architecture integrates a parameter-efficient MiniLM adaptation strategy to extract semantic representations from structured item-related educational metadata and a bidirectional recurrent encoder to model temporal learning patterns from behavioural logs. A gated fusion mechanism is then used to combine semantic and contextual information into learner representations, which are further integrated with collaborative user–item embeddings for top-K recommendation using a Bayesian personalised ranking objective. Experiments conducted on the EdNet-KT1 dataset under chronological splitting, full-corpus ranking, and fixed candidate-sampling protocols show that the collaborative-enhanced model achieves the highest-ranking performance among popularity-based, matrix factorisation, neural collaborative filtering, recurrent sequential, self-attention sequential, and ablation baselines. The model obtains an NDCG@10 of 0.1344 under full-corpus ranking and 0.5383 under candidate sampling, with statistically significant but practically modest improvements over the strongest baselines. Additional ablation, efficiency, and gate analyses indicate that semantic–contextual modelling is most effective when used as a residual enhancement to collaborative recommendation rather than as a standalone replacement. These results suggest that parameter-efficient semantic–sequential modelling, when combined with collaborative preference signals, offers a promising direction for scalable and evidence-based educational recommender systems. Full article

Gully erosion is a major driver of irreversible soil loss in Northeast China’s Mollisol belt, a region that supplies roughly one-quarter of the national grain output. Existing susceptibility assessments in this region have rarely combined multi-model comparison with spatially explicit cross-validation, and the predictive contribution of composite anthropogenic indicators such as the Human Footprint Index (HFI) has not been quantitatively benchmarked against conventional topographic variables. This study addresses these gaps for the Heihe region by combining an inventory of 4020 gully polygons supported by field checks in Xunke County, 16 VIF-screened environmental factors, three tree-based ensemble models and a logistic regression baseline. Under stratified random splitting, XGBoost achieved the highest discrimination (AUC = 0.95, $\kappa$ = 0.74); under leave-one-district-out spatial cross-validation all tree-based models retained AUC above 0.83, confirming that random-split metrics overestimate discrimination by approximately 0.11 AUC units due to spatial autocorrelation and inter-district covariate shift. SHAP analysis identified LULC and HFI as the dominant predictors, exceeding all topographic variables, while slope gradient contributed least—consistent with the low-relief, intensively cultivated character of the study area. Susceptibility was highest in the southwestern agricultural lowlands. A one-factor sensitivity test in which only NDVI was increased by 20% suggested a reduction in modelled high-susceptibility area of approximately 12%, although co-occurring land-cover and hydrological changes were not simulated. The multi-model framework, integrating spatial cross-validation and post hoc interpretability, provides an explicit estimate of conventional evaluation optimism and supports spatially differentiated erosion management. Full article

Little existing work addresses the joint design of time slot allocation and power splitting ratio optimization in simultaneous wireless information and power transfer (SWIPT)-enabled wireless sensor networks (WSNs). To fill this gap, this paper proposes a novel energy-sustainable framework termed ETAPS that co-optimizes time slot allocation and power splitting ratio for SWIPT-enabled WSNs. A dedicated frame structure is designed that partitions each cluster member (CM) into four operational modes for slot scheduling, toward conflict-free and coordinated resource allocation among CMs. A dynamic power splitting strategy is further developed to adaptively refine slot allocation for CMs and derive the optimal power splitting ratio for the cluster head (CH). Comprehensive numerical simulations are performed to validate the proposed scheme. The results demonstrate that ETAPS maintains effective energy sustainability even under limited energy input from the energy access point (EAP). When the EAP provides a sufficient energy supply, the optimal power splitting ratio converges to 0.9. Moreover, under sufficient transmit power at CMs, ETAPS adaptively allocates transmission time from CMs to the CH by setting the optimal power splitting ratio to 0.6. Full article

Steam methane reforming (SMR) powered by induction heating offers a promising route for low CO₂-emission hydrogen production, but predictive modelling remains challenging because the available experimental data are limited and heterogeneous. This study proposes a hybrid Bayesian neural network (H-BNN) to predict the mass of hydrogen (MoH) from literature-derived SMR data using operating variables including temperature, flow rate, power input, time-on-stream, and interval duration. Feedforward neural network (FNN) and classical Bayesian neural network (BNN) models were also developed as benchmarks, and all three architectures were evaluated with ReLU, Tanh, and GELU activation functions. To address data scarcity, only the training split was augmented at scales of $k=2$, 5, and 10, while the validation and test sets were kept unchanged. The H-BNN combines deterministic feature extraction with Bayesian uncertainty-aware prediction, enabling a balance between accuracy and uncertainty representation. Across the validation-selected models, test performance reached R² ∼ 0.9894 to 0.9969, with mean absolute errors of 0.0126 g to 0.0217 g. The strongest advantage appeared at k = 2, where the H-BNN outperformed the benchmark models. Overall, the proposed H-BNN is a promising framework for hydrogen prediction under data-scarce conditions, although its predictive intervals remain informative rather than fully calibrated. Full article

Efficient utilization of carbon dioxide (CO₂) is a critical route toward carbon cycling and low-carbon energy systems. Compared with conventional thermocatalysis, photocatalysis, and electrocatalysis, plasma catalysis can activate CO₂ under relatively mild conditions through high-energy electrons, vibrationally excited molecules, radicals, and other reactive species, while catalytic surfaces can redirect reaction pathways and improve selectivity. Rather than only compiling reported performances, this review critically evaluates plasma-catalytic CO₂-to-energy conversion from three perspectives: reliable mechanistic knowledge, unresolved uncertainties in plasma–catalyst synergy, and the practical credibility of reactor–catalyst combinations. The fundamentals of non-thermal plasma, CO₂ activation, key metrics, plasma–catalyst coupling, and catalyst/reactor/operation factors are first clarified. Representative advances in CO₂ splitting, CO₂ hydrogenation, dry reforming, and CO₂–H₂O co-conversion are then compared with attention to energy input, selectivity, power determination, and data comparability. Finally, the key barriers to industrial deployment are discussed, including low energy efficiency, long-term catalyst stability under plasma exposure, uncertain absorbed-power measurement, incomplete carbon/oxygen balances, scale-up of filamentary discharges, and the lack of standardized reporting protocols. This review aims to provide a critical reference for mechanism-guided catalyst design, reactor engineering, and realistic process assessment in plasma-catalytic CO₂ utilization. Full article

Faced with a global energy crisis and ecological degradation, overall water splitting (OWS) is a pivotal approach for renewable energy conversion and storage. However, its industrial application is hindered by the high energy barriers/sluggish kinetics of the anodic oxygen evolution reaction (OER), as well as the scarcity of precious metal catalysts limiting large-scale deployment. Herein, a cobalt-based layered double hydroxide (Co-LDH) was used as the precursor, and a multi-strategy synergistic modification (hydrothermal synthesis, Fe doping, sulfurization, and external magnetic field magnetization) was applied to fabricate the Fe-Co₃S₄-MS-20 min electrocatalyst. This strategy establishes Fe-Co bimetallic synergistic active centers, and magnetic treatment modulates the electron configuration of Fe 3d orbitals without changing the material’s lattice spacing or morphology. Structural characterizations and electrochemical measurements were used to investigate the effects of combined modifications on the catalyst’s phase structure, morphology, electronic structure, and trifunctional catalytic performance toward the hydrogen evolution reaction (HER), OER, and urea oxidation reaction (UOR). The Fe-Co₃S₄-MS-20 min catalyst exhibits a larger electrochemical active surface area, lower charge transfer resistance, and smaller Tafel slope in 1 M KOH, it achieves overpotentials of 165 mV for HER (10 mA·cm⁻²) and 310 mV for OER (100 mA·cm⁻²), along with superior UOR performance and long-term stability. In situ impedance and Raman spectroscopy confirm that magnetization accelerates charge transfer and promotes in situ reconstruction. Synergistic multi-strategy regulation optimizes the electronic structure of active centers, reducing electrocatalytic energy barriers. This work provides new insights into designing high-performance non-precious metal electrocatalysts and offers experimental support for external magnetic field regulation in electrocatalyst modification. Full article

An in-depth analysis of the drivers of agricultural emissions at the farm-gate level is crucial for achieving net-zero emissions by 2050. This study examines the short- and long-run effects of crop production on farm-gate emissions in the regional trading bloc (RTB) member states in Africa. Crop production was proxied by cereals, roots and tubers, vegetables, and fruits production, and emissions were split into methane (CH₄) and nitrous oxide (N₂O) emissions. Data on these variables were collected from 30 RTB member states from 1990 to 2022 and were analyzed using the cross-sectionally augmented autoregressive distributive lag approach. The pooled mean group was used as a robustness check, and a sensitivity analysis was conducted to ensure the reliability of the study findings. The results revealed that cereal production increases farm-gate CH₄ and N₂O emissions in the short and long run. The average increase ranges from 1.0021 to 1.0033 kilotons CO₂–eq yr⁻¹ for CH₄, and from 1.0024 to 1.0035 kilotons CO₂–eq yr⁻¹ for N₂O. In addition, fruit production increases farm-gate CH₄ emissions by an average of 1.0023 kiloton CO₂–eq yr⁻¹ in the long run. Thus, cereal production has a more adverse effect on N₂O than CH₄ emissions, while the opposite is true for fruit production in the RTB member states’ Nationally Determined Contributions. With respect to mediation, cropland expansion (proxied by area harvested) plays a partial intermediary role in the impact of crop production on farm-gate CH₄ and N₂O emissions in the short run and CH₄ emissions in the long run. However, it assumes a full mediation role in the long run and has an effect on crop production in farm-gate N₂O emissions. Therefore, targeted use of nitrogen fertilizer and crop rotations to reduce cereal-related N₂O and CH₄ emissions, respectively, would be viable strategies. The use of a drip irrigation system in fruit production to reduce CH₄ emissions and the scaling up of secured subsidies should also be explored. Finally, these recommendations should be incorporated into the Africa’s RTB member states’ Nationally Determined Contributions and the African Union’s Agenda 2063. Full article

Land surface temperature (LST) is a core physical parameter that characterizes land surface processes and surface-atmosphere energy exchange. As the demand for high-accuracy LST products intensifies across diverse research domains—including climate science, hydrology, and ecosystem modeling—the systematic quantification of pixel-level retrieval uncertainties has become essential for generating long-term, consistent Climate Data Records (CDRs). However, existing studies predominantly emphasize algorithmic development or localized validation, with limited attention to systematic cross-site and long-term uncertainty assessments. This gap impedes a comprehensive understanding of the compositional structure and spatiotemporal variability of LST retrieval uncertainties under heterogeneous surface and atmospheric conditions. In this study, based on the improved generalized split-window (GSW) algorithm and error propagation theory, the total uncertainty (U_total) and its four primary components—algorithm uncertainty (U_a), land surface emissivity uncertainty (U_e), noise equivalent delta temperature uncertainty (U_n), and atmospheric water vapor uncertainty (U_w)—at the pixel level over long time series and across multiple sites are quantified. Our analysis spans a 21-year period (2003–2023) and encompasses multiple geographically distributed sites, utilizing high-quality Moderate Resolution Imaging Spectroradiometer (MODIS) thermal infrared data—specifically MYD11_L2 and MOD11_L2 products—collocated at the locations of 15 globally distributed ground-based reference sites. These sites are used to represent diverse climatic regimes and land-cover conditions, rather than to provide point-scale “true” LST values for residual-based validation. Results show that the interquartile range (IQR) of U_total is consistently concentrated between 1.0 and 1.2 K, demonstrating long-term stability. Systematic differences in U_total are identified across sensor platforms and diurnal cycles: U_total for Aqua/MYD data (1.13–1.25 K) is marginally higher than that for Terra/MOD data (1.05–1.17 K); similarly, daytime U_total (1.08–1.23 K) is generally slightly elevated relative to nighttime U_total (1.05–1.18 K). The contributions of individual uncertainty components to U_total exhibit substantial variation, with mean relative contributions of 81.97%, 11.32%, 4.46%, and 2.25% for U_e, U_a, U_n, and U_w, respectively. The dominant drivers of U_total differ markedly across climatic regions: in arid regions, U_total is predominantly governed by U_e, termed “emissivity-dominated,” accounting for over 85% of the total; conversely, humid tropical regions exhibit a “surface-atmosphere co-influenced” regime, characterized by a reduced contribution from U_e and correspondingly enhanced contributions from U_a and U_w. Furthermore, U_total decreases with increasing total column water vapor (TCWV) (Pearson correlation coefficient r = −0.498; linear slope k = −0.0425 K/(g/cm²)), and increases with increasing viewing zenith angle (VZA) (r = 0.208; k = 0.0022 K/degree). While U_a, U_n, and U_w all increase with TCWV, U_e decreases. Full article

Calcium copper titanate (CaCu₃Ti₄O₁₂, abbreviated as CCTO) has emerged as a versatile, high-performance material distinguished by its remarkable dielectric, photocatalytic, and environmental properties, positioning it at the forefront of ongoing research and technological innovation. This review provides a comprehensive analysis of CCTO, emphasizing its growing relevance in catalytic and environmental applications. Beginning with an overview of its unique structural and dielectric properties, we discuss how these attributes underpin CCTO’s multifunctionality. Various synthesis methods are examined for their effects on CCTO’s microstructure and performance. Furthermore, we investigate the photocatalytic potential of CCTO under visible light, particularly for applications such as water splitting, CO₂ reduction, and degradation of organic pollutants. Environmental applications, including gas sensing and wastewater treatment, are also evaluated, highlighting CCTO’s chemical robustness and suitability under diverse operating conditions. Lastly, key challenges in scalability, cost, and environmental adaptability are discussed, along with future directions, including hybrid composite development and machine-learning-assisted material design. Together, these insights position CCTO as a promising material for advancing sustainable technologies in energy and the environment. Full article