Browse Articles
Discover research articles across all indexed journals
Dual-adaptive imputation graph neural network for knowledge-aware recommendation
Abstract Recommender systems play a vital role in enhancing user experience by efficiently delivering personalized and relevant content. While knowledge graph-based recommender systems effectively alleviate the data sparsity and cold-start challenges of traditional approaches, they still suffer from two major limitations: (1) insufficient utilization of the user-item interaction matrix and (2) suboptimal integration of heterogeneous knowledge graph signals with collaborative information. In this work, we propose DAIGNN (Dual-Adaptive Imputation Graph Neural Network), a novel recommendation framework designed to overcome these limitations through three key innovations. First, we introduce a similarity-driven imputation mechanism that constructs an Imputation Graph using pseudo-ratings, thereby enhancing graph connectivity and significantly reducing data sparsity. Second, we incorporate multiple auxiliary information sources on both the user and item sides, enabling DAIGNN to capture richer contextual and relational semantics beyond conventional user-item interactions. Third, we develop a dual-adaptive feature fusion mechanism that learns optimal fusion weights to dynamically integrate heterogeneous information from multiple graph sources. Extensive experiments conducted on four real-world datasets demonstrate the superior effectiveness of DAIGNN. On average, it achieves a 3.1% improvement in AUC and a 2.0% improvement in F1-score over state-of-the-art baselines, confirming its robustness across diverse settings.
Abiotic synthesis of RNase-resistant phosphodiester and pyrophosphate-linked polymers via thermodynamically controlled wet-dry cycling
Biotype dynamics of the common house mosquito Culex pipiens in the anthropogenic environment in winter: a citizen science approach
Abstract The two biotypes of the mosquito species Culex pipiens s.s. (biotype pipiens and biotype molestus ) differ in overwintering strategy. Biotype pipiens enters diapause, whereas biotype molestus remains active year-round. In the Netherlands, flooded crawlspaces may provide sheltered development sites for biotype molestus in winter. We therefore investigated the effects of human population density and crawlspace flooding on the presence of Cx. pipiens s.s. biotypes and their hybrids using citizen-submitted mosquito samples via the platform ‘Muggenradar.’ Mosquitoes were categorized by population density and crawlspace condition. In addition, biotypes and feeding patterns were identified using molecular analyses. Descriptive analyses indicated that biotype molestus was more abundant in areas with high population density (24.1% at ’low’ versus 58.6% at ‘high’), while biotype pipiens dominated in areas of low population density (19.5% at ‘high’ versus 63.3% at ‘low’). Similarly, we found an association of biotype molestus with flooded crawlspaces (68.4% of specimens). However, these patterns were not fully supported by a multivariate analysis. Blood meal analyses revealed that nearly all biotype molestus (98.2%) and hybrid mosquitoes (100%) fed on humans. Surprisingly, six blood-engorged biotype pipiens were collected. We conclude that anthropogenic environments may influence the occurrence of Cx. pipiens s.s. biotypes and hybrids in winter.
An Illumina-based amplicon sequencing approach designed to determine grapevine fanleaf virus isolates
Evaluating hygienic practices in cattle slaughterhouses and beef retail shops in Sidama region, Ethiopia: implications for public health
Osteogenic and antibacterial effects of double antibiotic-loaded microspheres
Unveiling the quality profiles of honey from selected districts of South Wollo, Ethiopia
Enhanced degradation and defluorination of perfluorooctane sulfonate (PFOS) in tap water using gas-dispersed cold atmospheric plasma
Abstract Per- and polyfluoroalkyl substances (PFAS) are extremely persistent contaminants owing to the exceptional chemical stability of carbon–fluorine (C–F) bonds. Consequently, conventional wastewater treatments are largely ineffective, as they capture but fail to destroy PFAS, leading to the accumulation of concentrated wastes. In this study, we demonstrate that gas dispersion-assisted cold atmospheric plasma (CAP) enables rapid degradation and partial defluorination of perfluorooctane sulfonate (PFOS) in tap water. Operating under ambient conditions, CAP generates a rich mixture of oxidative and reductive reactive species, including solvated electrons and hydroxyl radicals, which are proposed to contribute to PFOS degradation and defluorination. Air gas dispersion enhances hydrodynamic mixing and enriches PFOS at the plasma-liquid interface, promoting interfacial microdischarges and concentrated short-lived reactive species that enhance oxidative and reductive degradation pathways. At high PFOS concentrations in tap water, gas dispersion-assisted CAP achieved 99.99% PFOS degradation with partial defluorination of 35%. With gas dispersion, degradation followed apparent first-order kinetics, with a rate constant of 0.42 1/min and a half-life of 1.6 min. In both conditions, with and without gas dispersion, analysis of measured transformation products (TPs) revealed stepwise degradation pathways of PFOS, with fluorine mass balance recoveries ranging from 31 to 106%. The lowest electrical energy per order ( EEO ) achieved was 39 kWh/m 3 /order. These results demonstrate the efficient degradation of PFOS, while the measured fluoride ion release confirms partial defluorination, highlighting gas dispersed CAP as a promising chemical-free and energy-efficient technology for PFAS remediation in water systems.
The relationship between perceived parental knowledge and adolescent gambling
First report of tenacibaculosis in wild-caught Pacific salmon
MAE-YOLO improves small object detection for intelligent inspection
Abstract Intelligent inspection technology has become increasingly popular in industrial fields such as power facility maintenance (e.g., identifying cracked insulators, corroded transformers), traffic management (e.g., detecting vehicle anomalies), and industrial equipment upkeep (e.g., spotting surface defects on machinery, verifying the presence of small components like bolts and valves). The targets in these scenarios are often small-sized, well-defined objects that are critical to operational safety and efficiency. Although the traditional deep learning methods such as YOLO series have made progress in this field, however, in the detection of small targets under complex background, it still suffers from high false detection and miss rates, as well as high computational complexity, making it difficult to meet the real-time requirements of practical applications. In this article, an intelligent inspection model named MAE-YOLO is proposed. Firstly, a multi-scale edge space feature extraction module is proposed to optimize the edge and space feature extraction, which significantly improves the detection accuracy of small targets. Meanwhile, the adaptive multi-scale context fusion network is introduced to integrate the features of different scales effectively, thereby enhancing the robustness and adaptability of the model in dynamic environments. Finally, we propose an adaptive cavity shared detection head to further reduce the false detection and missing detection rate in multi-scale detection. By synergistically integrating MSESTE for edge-aware feature extraction, AMCFN for adaptive multi-scale context fusion, and ELCIN for parameter-efficient detection, MAE-YOLO achieves both lightweight design and high accuracy for small objects. Experimental results on the VisDrone2019 dataset show that the accuracy of MAE-YOLO is improved by 2.6% compared to the original YOLOv8n model. According to the results of the self-collected dataset, it can be seen that MAE-YOLO only needs 4.7 MB, which is reduced by 24% compared with YOLOv8n, while maintaining high detection accuracy. Unlike existing lightweight methods that often sacrifice edge details for efficiency, MAE-YOLO preserves fine-grained object boundaries through Sobel-based edge enhancement while reducing detection head parameters by 53%, achieving a superior accuracy-size trade-off (35.1% mAP@50 at 4.57 MB) compared to recent detectors such as SOD-YOLO (30.08% mAP@50). To facilitate reproducibility and further research, the source code of MAE-YOLO has been released at: https://github.com/970334745/MAE-YOLO .
EEG blink and gaze control using random forest classification for accessible assistive robotic navigation in real world conditions
Study on the evolutionary mechanism and synergistic control technology of floor heave in deep high-stress roadways
Calibration-free physics-informed multi-task residual U-Net for simultaneous denoising and gas pressure retrieval from noisy voigt spectra
Abstract We present a machine learning framework based on a one-dimensional U-Net (1D U-Net) that simultaneously performs spectral denoising and pressure estimation within a unified architecture. The model is trained on simulated Voigt profiles of the P(21) CO absorption line over pressures ranging from of 1 mbar to 2 bar. To ensure realistic conditions, simulated spectra are superimposed with experimentally captured noise, making them nearly indistinguishable from real experimental scans and forcing the model to recover clean spectra from noisy inputs. Quantitative assessments indicate excellent reconstruction performance, with a Pearson correlation coefficient (PCC) approaching unity, a signal-to-noise ratio (SNR) exceeding 35 dB, and both mean absolute error (MAE) and mean squared error (MSE) remaining close to zero. The model accurately predicts pressures for 200 unseen spectra using only spectral features, bypassing traditional linewidth analysis. At 1.25 bar, the U-Net yields virtually zero error (AE ≈ 0, SE ≈ 0), demonstrating sub-percent deviation and excellent consistency with the simulated reference. Experimental validation on a difference-frequency generation (DFG) spectrometer confirms robust performance, with ≈ 70% of traces reaching a peak SNR (PSNR) above 34 dB. Pressure estimation from ramp-based scans further demonstrates high accuracy, achieving minimal errors at 539.1 mbar (AEP = 0.002, SEP = 4.0 × 10⁻⁶). These findings establish the 1D U-Net as an efficient and reliable alternative to conventional noise-reduction and pressure-estimation techniques, simplifying mid-infrared spectroscopy workflows while ensuring high fidelity and stability.
Thermo rheological performance of DFNS enhanced asphalt binder modified with waste cooking oil and waste rubber powder
Abstract This study investigates the combined effects of dendritic fibrous nanosilica (DFNS), waste cooking oil (WCO), and waste rubber powder (WRP) as multifunctional modifiers for asphalt binders. Seven asphalt formulations were prepared and evaluated through conventional tests, dynamic shear rheometry (DSR), bending beam rheometry (BBR), fatigue time sweep analysis, and thermal characterization. Results showed that the DFNS/WCO/WRP modified binder exhibited improved rheological and mechanical performance compared with the base binder. For example, the rutting factor (G*/sinδ) increased from 2.18 to 3.10 kPa at 52 °C, while creep stiffness at − 24 °C decreased from 351.2 to 301.3 MPa, indicating enhanced resistance to low temperature cracking. Fatigue testing also demonstrated improved durability, with the normalized modulus remaining 0.756 after 10,000 s, compared with 0.602 for the base binder. These improvements are attributed to the synergistic interaction of DFNS providing structural reinforcement, WRP contributing elastic recovery, and WCO enhancing flexibility and dispersion within the binder matrix. The results demonstrate that the DFNS/WCO/WRP system is a promising modification strategy for improving the rheological performance and durability of asphalt binders.
Integrated multi-scale aeromagnetic, gravity, and remote-sensing analysis for mapping basement fabric and structural architecture in the ِِِAswan region, Southern Egypt
Abstract An integrated imaging workflow combining aeromagnetic, gravity, optical, and radar satellite datasets was applied to characterize basement structures beneath the Aswan area, southern Egypt. Optical and radar satellite data lineament analyses, together with Digital Elevation Model (DEM) derived hillshades, and automated lineament extraction revealed a dominant NW–SE orientation related to the Pan‑African shear fabric with minor E–W unloading joints, and reactivated NE–SW shear trends. Bouguer gravity anomalies (–48.3 to − 13.5 mGal) clustered spatially and are characterized by higher values (–13.5 to − 22.0 mGal) near Aswan, aligned with dense crystalline horsts, whereas lower anomalies (–35.5 to − 48.3 mGal) delineated sediment-filled grabens, dominant along the west and the southeast. High-pass filtering highlights shallow N-S lineaments parallel to the Nile valley and NE-SW fractures. In contrast, low-pass filtering mapped the broader basement geometry, showing a gentle NW-SE to E-W slope with uplifted shoulders east of the study area. Two-dimensional magnetic modeling, supported by Euler deconvolution, revealed significant basement depth variation across the study area, ranging from about 350–400 m in shallow zones to 1,800–2,600 m in deeper sectors. These integrated imaging results revealed horst–graben structures, identified favorable targets for groundwater and mineral exploration, guided infrastructure planning, and demonstrated the value of integrated geophysical workflows.