Search Results (208)

Background/Objectives: Depression, anxiety disorders, and substance use disorders frequently coexist in clinical settings and are main factors that worsen a patient’s prognosis. However, traditional artificial intelligence-based mental health studies have limitations in capturing the complex symptoms that occur in actual counseling situations by relying on social media data or focusing on binomial classification of single diseases. This study proposes a multi-label classification model that simultaneously detects the coexistence of depression, anxiety, and substance use disorder in actual counseling dialogue texts, and applies the Shapley Additive Explanatory (SHAP) method to explain the clinical basis of model prediction. Methods: We retrospectively analyzed 1661 de-identified Korean-language counseling session transcripts obtained from the publicly available AI Hub “Mental Health Counseling Dialogue” dataset (Republic of Korea; sessions collected between 2021 and 2023 from accredited domestic mental health counseling centers). Each session averaged 30 min (≈5000 Korean characters). Labeling was performed by two licensed clinical psychologists (inter-rater Cohen’s κ = 0.82). A Hierarchical Attention Network with Bidirectional LSTM (HAN-BiLSTM) was constructed; performance was compared with six baselines (Flat LSTM, TextCNN, KR-BERT, KoBERT, KoELECTRA, KLUE-RoBERTa) using stratified 5-fold cross-validation, paired t-tests with Bonferroni correction, and McNemar’s test. Top-ranked SHAP tokens were independently rated for clinical face validity by three psychiatrists. Results: The proposed model outperformed the baseline model not only for the labels of depression (F1 = 0.90) and anxiety (F1 = 0.85) but also for substance use disorder (F1 = 0.78) with poor data, achieving a macro-averaged F1 of 0.84 (95% CI 0.82–0.86; all p < 0.001 versus baselines). As a result of the SHAP analysis, clinically significant keywords such as “I want to die,” “anxiety,” and “drink” were identified as the model’s main basis for judgment, accurately tracking the client’s state, which dynamically changed as the dialogue progressed; three independent psychiatrists rated 88.7% of the top-15 SHAP tokens per label as clinically meaningful (Fleiss’s κ = 0.76). Conclusions: This study demonstrated that a deep learning-based multi-label approach is effective in early screening of complex mental health problems. In particular, the introduction of explainable AI (XAI) increases clinicians’ trust and suggests that it can be used as an AI-based clinical decision support system (CDSS) in the future. Full article

Driven by the digital transformation of territorial spatial governance, traditional urban planning is irreversibly shifting towards a data-driven empirical paradigm. However, constrained by mimetic isomorphism and path dependence, many geography-based regional universities remain trapped in an educational dilemma: they overemphasize morphological representation while marginalizing quantitative decision-making, fostering a structural mismatch between graduate competencies and industry demands. To explore a systematic pathway out of this dilemma, this study chronicles a three-year pedagogical intervention utilizing a mixed-methods design with a historical control cohort (N = 275) within the urban planning program of Gansu Agricultural University—a regional institution situated in a less-developed frontier where territorial renewal demands macro-spatial synthesis over aesthetic forms. The intervention strategically redefined the graduate competency profile as “spatial data analysts”, constructing a pedagogical model comprising foundational algorithmic training, cross-disciplinary faculty collaboration, and real-world Project-Based Learning (PBL), coupled with a restructured, evidence-based evaluation system. Longitudinal tracking and quantitative analyses indicate a structural alignment with elevated educational efficacy. At the macro level of employment trajectories, the proportion of graduates securing knowledge-intensive data positions experienced a structural shift, rising from a baseline of 14.5% to 42.5%, reflecting an enhanced capacity to capitalize on expanding societal demands. At the meso level of practical competence, the award rate in high-level professional competitions increased by 35.4%. At the micro cognitive level, the new evaluation mechanism is associated with a successful redirection of students’ cognitive resources toward algorithmic logic and policy translation (p < 0.001) while highly significantly enhancing their self-efficacy in tackling complex, wicked engineering problems (p < 0.001). Rather than isolating pure causal mechanics, this study interprets these systemic gains as a contextual realignment of academic supply. It provides a context-sensitive, reproducible methodological reference for cultivating professional distinctiveness and reshaping the spatial planning education system in the digital era. Full article

In cardiovascular interventional procedures, real-time, precise monitoring of minute strain and temperature fluctuations at the catheter tip is essential to improving both the safety and efficacy of these interventions. Fiber Bragg grating (FBG)-based sensors present a promising solution owing to their diminutive size and immunity to electromagnetic interference; however, the inherent cross-sensitivity between strain and temperature remains a significant obstacle. This paper introduces a dual-FBG fiber optic sensing structure that leverages machine learning techniques. The system incorporates two FBGs: one set acts as the primary sensing element, positioned within a simulated catheter and affixed to the substrate under examination with cyanoacrylate adhesive to detect composite strain and temperature signals; the second set is spirally wound around the catheter surface to solely measure temperature, thus effectively isolating temperature interference. Additionally, a machine learning model is employed to learn the nonlinear mapping between the recorded FBG spectra and the actual strain and temperature parameters. Experimental validation was conducted within the physiologically relevant temperature range of 20 °C to 45 °C. The findings indicate that the proposed machine learning model can successfully decouple strain and temperature, achieving high-precision predictions even in situations where the sensing unit exhibits a slight nonlinear response due to adhesive bonding. This study substantiates the feasibility of utilizing machine learning-enhanced dual-FBG structures for multi-parameter sensing in complex environments. The proposed methodology presents a promising avenue for the development of next-generation smart optical fiber sensors intended for application in catheter systems. Full article

Vessel traffic service systems in remote or newly established maritime regions face significant operational limitations due to the scarcity of historical AIS data, which undermines the reliability of trajectory-based situational awareness and collision risk assessment. Existing deep learning models, predominantly validated on data-rich major shipping corridors, suffer severe performance degradation under cross-domain deployment, rendering them impractical for vessel traffic management in underserved waters. To bridge this operational gap, this study proposes a Boundary-Aware Distillation and LoRA-Based Transfer (BD-LT) framework that enables reliable trajectory prediction with as few as 132 target-domain trajectories. The framework integrates HDBSCAN-based geographic-semantic domain partitioning, a Time-Aware Transformer with Time2Vec encoding for irregular AIS sampling, hybrid knowledge distillation with error-boundary gating for selective cross-domain transfer, and LoRA-based parameter-efficient adaptation to mitigate overfitting. Validated on NOAA full-scale AIS measurements, the framework reduces the 60 min Final Displacement Error by 35.2% relative to the no-framework baseline, consistently outperforming state-of-the-art models across all prediction horizons, with statistical reliability confirmed via bootstrap resampling. These results demonstrate the practical feasibility of deploying data-driven trajectory prediction in maritime regions where conventional approaches have historically been inapplicable, with direct implications for collision avoidance decision support and port approach traffic management. Full article

The remote tower operation mode has changed how controllers acquire, integrate, and interpret operational information, making Situation Awareness (SA) prediction more challenging because of the coupling of multiple heterogeneous information sources. To address the limitations of existing physiological-data-based studies in modeling cross-modal relationships and deep multimodal interactions, this study proposes MGRF-Net, a multimodal physiological data-driven model for predicting remote tower controllers’ SA. The model first encodes eye-tracking, electroencephalography, electrocardiography, and electrodermal activity signals independently to obtain high-level temporal representations. A graph attention-enhanced relational learning module is then introduced to capture interactive dependencies among modalities, followed by a dual-branch gated fusion mechanism to adaptively integrate multimodal information and improve prediction stability. Using multimodal physiological data collected from a remote tower simulation experiment and evaluated with 12-fold cross-validation, MGRF-Net achieved 0.0658 RMSE, 0.0461 MAE, 0.8579 R², and 0.9308 PCC, outperforming LightGBM, MLP, PatchTST, iTransformer, and TimeMixer. Ablation experiments and SHAP analysis further confirmed the effectiveness and interpretability of the proposed model. The results indicate that MGRF-Net can effectively capture cross-modal coupling patterns in the formation of controllers’ SA and provides a promising approach for complex cognitive state monitoring and intelligent assistance in remote tower operations. Full article

Higher education institutions have been identified as playing a pivotal role in the advancement of the Sustainable Development Goals (SDGs) through teaching, research, and professional training. However, previous studies indicate that the integration of the SDGs into students’ learning processes remains uneven, with students often demonstrating general awareness but limited structured knowledge of the 2030 Agenda. This exploratory case-based study examines how SDGs are encountered, understood, and appropriated by university students at Universidad Latina de Costa Rica within a 2024–2025 collaborative project with the National System of Accreditation of Higher Education (SINAES). A mixed-method, cross-sectional design was employed using a structured questionnaire administered to 434 students across campuses and disciplines. The study analyzes the relationship between exposure, knowledge, and disciplinary appropriation. Findings show that while institutional visibility of the SDGs is associated with greater conceptual familiarity (r = 0.85; p < 0.01), this does not necessarily translate into their integration within disciplinary training. Students tend to interpret the SDGs primarily as ethical frameworks, with limited connection to professional practice. These results suggest that the main challenge lies not in awareness, but in curricular and pedagogical integration, highlighting the need to approach SDG implementation as a situated educational process shaped by disciplinary contexts and learning environments. Full article

Orbital maneuver detection is a core component of space situational awareness. The multi-scale characteristics of satellite orbital behavior and sample imbalance issues lead to challenges in existing methods, including insufficient feature adaptation and limited detection accuracy. This paper proposes an Adaptive Spiking Gating Multi-Scale Liquid State Machine (ASG-MSLSM) for orbital maneuver detection based on variations in satellite orbital parameters. The method integrates multi-scale reservoir pools with different scale-dependent decay factors and Leaky Integrate-and-Fire (LIF) neurons to enhance multi-scale temporal feature extraction capability. A spiking gating network is designed to adaptively learn fusion weights for multi-scale features, replacing traditional fixed equal-weight fusion strategies. During training, weighted binary cross-entropy loss is employed to address class imbalance. Experimental results based on real satellite data demonstrate that the proposed method significantly outperforms baseline models in maneuver detection metrics, achieving higher recall, improving feature separability, and reducing both missed detections and false alarms. These results indicate that the proposed method provides a robust solution for orbital maneuver detection. Full article

This study applies a dynamic strategic foresight to examine how Unmanned Aerial Systems (UAS)-based Advanced Air Mobility (AAM), supported by Advanced Communication Systems (ACS), can be integrated into a coherent System of Systems (SoS) for sustainable and effective Disaster Management (DM). These three domains (AAM, ACS, and DM) form a strongly coupled Internet of Things (IoT) triad within an integrated SoS. Using lessons learned from previous or running research projects of the contributing authors, i.e., SUDEM, REGUAS, 5G!Drones, and ETHER, the foresight identifies key enablers—including resilient 5G/6G communication architectures, interoperable data fusion frameworks, and UAS-supported situational awareness. It highlights structural challenges such as fragmented standards, limited cross-agency data integration, and gaps in ACS redundancy for emergency operations. The resulting roadmap outlines development priorities for ACS-enabled AAM, from unified communication protocols and hybrid TN-NTN architectures to education and capacity-building for digital-centric DM. Practically, the findings suggest that policymakers should prioritise harmonised regulatory frameworks for AAM-ACS interoperability and invest in global data exchange standards, while system designers should incorporate redundant communication layers and modular SoS architectures to ensure operational continuity under extreme conditions. Full article

For Unmanned Aerial Vehicles (UAVs), optical sensing for reliable landing and the detection of the landing area is a crucial element. In low-light conditions, at night, and in foggy weather, where optical sensing is not feasible, thermal imaging can be utilised. Although this situation has been widely researched, most UAV landing approaches rely on GNSS assistance or single-mode detection, which limits their robustness and scalability in real-world operations. This study proposes an actively heated thermal helicopter landing pad designed using electrically powered resistive heating elements and a high-emissivity surface coating. Furthermore, optical and thermal images collected during actual UAV flight experiments under daytime and night-time conditions were processed using image fusion techniques with AVGF, DWTF, GPF, LPF, MPF, and HWTF fusions, and their performance in deep learning models was compared. The obtained optical, thermal, and fused datasets are used to train and evaluate deep learning-based helicopter landing pad detection models based on the YOLOv8 architecture. Experimental results show that models trained with single-mode data exhibit limited cross-domain generalisation, while fusion-based learning significantly improves detection robustness in optical and thermal domains. Among the evaluated methods, LPF, MPF and HWTF provide the most consistent performance improvements. The findings indicate that electrically heated thermal helicopter landing pads, when combined with image fusion and deep learning-based detection, can increase the landing detectability of UAVs at night and in low-visibility conditions. This detection-focused approach contributes to UAV flight safety by enhancing the visibility of the landing area without relying on active infrared markers or additional navigation infrastructure. Full article

Digital transformation in higher education increasingly hinges on the robustness and governability of Information and Communication Technology (ICT) infrastructures, with campus Wi-Fi networks serving as the operational backbone of digital learning, research collaboration, and administrative services. In large universities, these networks typically evolve into heterogeneous, multi-vendor environments, introducing ongoing challenges in monitoring coherence, configuration governance, and cross-platform performance diagnosis. Despite the centrality of these issues, smart campus scholarship has paid limited attention to day-to-day operational management. This study examines the design and operational performance of a dual-platform Wi-Fi network management architecture implemented at Mahasarakham University, Thailand. The architecture strategically integrates SolarWinds and LibreNMS to combine centralized network-wide visibility with fine-grained, device-level diagnostics across a multi-vendor infrastructure. An engineering-oriented mixed-method approach was employed, drawing on production monitoring logs and semi-structured interviews with campus network engineers. Findings indicate that SolarWinds strengthens configuration oversight and campus-level situational awareness, whereas LibreNMS enhances detailed performance analytics and accelerates fault isolation. Their coordinated deployment improves operational stability, diagnostic clarity, and long-term maintainability of campus Wi-Fi systems. The study provides practical architectural guidance for managing heterogeneous ICT infrastructures in smart campus and enterprise-scale environments. Full article

The Vietnam Mekong River Delta (VMRD) is a climate-sensitive region characterized by diverse ecosystems, including extensive mangrove forests that protect against sea-level rise and contribute to global carbon sequestration. Accurate land cover classification in the VMRD is essential but remains challenging due to complex landscapes and dynamic environmental conditions. The primary objective of this study is to propose a semi-supervised deep learning framework that integrates satellite indices with multi-temporal remote sensing data to address key classification challenges, particularly in situations where ground truth data is limited, as compared to unsupervised and supervised machine learning methods. Our comparative analysis across different sample sizes (500 to 6000 ground-truth data points) reveals critical insights into model performance and scalability. Supervised models, including Random Forest (RF), Support Vector Machine (SVM), and Convolutional Neural Network (CNN), demonstrated strong performance when sufficient labeled data were available, with CNN achieving the highest accuracy (0.97 at 6000 samples). However, at minimal sample sizes (500 sample points), these supervised approaches exhibited substantial limitations, with accuracies dropping dramatically (RF: 0.75, SVM: 0.80, CNN: 0.81). Supervised models also showed overfitting tendencies compared to official land cover statistics. In contrast, the semi-supervised approach (SoC4SS-FGVC) achieves remarkably high performance at small sample sizes (0.92 accuracy with 500 sample points), demonstrating strength under minimal data availability. The framework also showed improved capability in distinguishing spectrally similar land-cover classes and detecting environmentally sensitive types such as mangrove forests. Cross-validation with official statistics confirmed the semi-supervised model’s superior effectiveness in delineating paddy rice fields and its resistance to overfitting. The performance analysis demonstrates that SoC4SS-FGVC provides a practical and cost-effective solution for land cover mapping, particularly in regions where extensive ground-truth data collection is prohibitively expensive or logistically challenging. Full article

This article focuses on the origin of “Wenqi Theory”—a core domain of ancient Chinese literary theory—specifically Cao Pi’s proposition that “literature is governed by qi”. It situates this concept within the 21st-century context of cultural globalization to engage in dialogue with Western aesthetics, aiming to revitalize the theory through mutual learning between Chinese and Western civilizations and integrate it into the system of modern transformation for classical literary theory. From the perspective of contemporary theoretical reconstruction, the paper analyzes the modern discourse paradigm of “Wenqi Theory”, traces its philosophical roots, and points out that the “clearness” or “murkiness” of “Wenqi” directly influences the aesthetic value of writing and the evaluation of objects. The study reveals that “Wenqi Theory” possesses rich connotations and unifies multiple dialectical relationships such as author and text, macrocosm and microcosm, personal temperament and acquired cultivation, content and form, fully embodying the distinctive integration of Chinese cultural tradition. Furthermore, the paper studies the lineage of life aesthetics from “Qi-Theory” in philosophy and science to “Wenqi Theory” in literary criticism, and its importance in constructing modern discourse paradigms. Meanwhile, by utilizing the categories of “the sublime” and “the beautiful” in Western aesthetics, it reactivates the contemporary aesthetic implications of “Wenqi Theory” within the context of globalization and cross-cultural exchange. The article endeavours to place this seemingly esoteric concept of classical Chinese literary theory within a cross-cultural and cross-disciplinary philosophical horizon for systematic and theoretical interpretation, revealing its universal aesthetic value that transcends specific cultural backgrounds, thereby providing a possible paradigm for the modernization of traditional Chinese literary theory and its participation in international academic dialogue. Full article

The growing complexity and interdependence of critical infrastructures (CIs), increasingly exposed to natural and technological hazards, call for educational approaches to enhance resilience and risk management. This study examines trends, patterns, and challenges in integrating digital and immersive technologies into education and training for stakeholders in critical infrastructure management. A systematic review of peer-reviewed literature was conducted using Scopus as the primary source, covering the last decade and analyzing the corpus across six dimensions: technological approach, pedagogical model, hazard typology, infrastructure domain, stakeholder category, and implementation phase. Following the PRISMA framework, 5635 records were identified and screened through a multistage process combining rule-based filtering and manual review, resulting in 105 papers meeting the inclusion criteria. The analysis reveals a shift from classroom instruction and physical drills toward immersive, simulation-based, and data-informed learning ecosystems that strengthen situational awareness, procedural accuracy, and decision-making under stress. However, the review identifies persistent gaps in evaluation metrics, cross-sector frameworks, and collaborative learning environments that limit adoption. The findings underscore that digital and immersive technologies can reconfigure education and training frameworks, enabling the formation of Resilient Operators endowed with adaptive cognition, continuous learning capacities, and responsiveness to natural hazard-induced technological risks. Full article

Epistemic agency is increasingly recognised as an important focus in teacher education, yet little is known about how it is enacted when pre-service teachers (PSTs) reflect on learner struggles in Physical Sciences. This qualitative case study investigated how Physical Sciences PSTs enacted epistemic agency through reflective engagement with learner struggles. Situated in a final-year Classroom Research in Physical Sciences module, the study involved 73 PSTs’ written reflections and follow-up semi-structured interviews with eight participants. Data were analysed thematically, guided by the lens of epistemic agency. Findings revealed enactments of critical noticing, representational critique, and responsible pedagogical reasoning, with teacher noticing as a cross-cutting mechanism. The study concludes that structured reflection fosters epistemic virtues and recommends explicit integration of learning struggles and noticing practices in teacher education. Full article

Object counting in remote sensing images is valuable for applications such as urban planning and environmental monitoring. However, it remains challenging due to heterogeneous annotations, semantic ambiguity in open-vocabulary queries, and performance degradation of small targets. To address these limitations, we propose DR-CLIP (Deformable Remote CLIP), a vision–language model for remote sensing image counting that incorporates deformable visual feature extraction with text-guided prediction. DR-CLIP includes a (1) Region-to-Instruction (R2I) mechanism to convert points, bounding boxes, and polygons into a unified image–text training representation, a (2) Multi-scale Deformable Attention (MSDA) to enhance discriminative feature extraction across extreme scale variations and cluttered backgrounds, and a (3) Text-Guided Counting Head that establishes robust cross-modal alignment through contrastive learning, achieving open-vocabulary counting capability without category-specific retraining. On DOTA-v2.0, DR-CLIP achieves a Mean Absolute Error (MAE) of 2.34 and a Root Mean Squared Error (RMSE) of 3.89, outperforming baselines by 19.0% in MAE. The MSDA module significantly increases Small-Object Recall (SOR) to 0.824, which is especially effective in situations involving dense and small object counting. In cross-modal retrieval, DR-CLIP attains R@1 scores of 68.3% (image-to-text) and 72.1% (text-to-image) on the Remote Sensing Image Captioning Dataset (RSICD). The framework generalizes robustly, with only 8.7% performance degradation in cross-domain tests, which is significantly lower than the 23.4% drop observed in baseline methods. Full article