Search Results (1,747)

Amid the accelerated global transition toward sustainable electromobility, Chinese Electric Vehicles (EVs) have forged a complex, evolving communication ecosystem across overseas social media platforms. Conceptualizing global digital discourse as a complex adaptive system (CAS), this study integrates CAS theory with Competitive Framing theory to systematically elucidate the thematic configurations, framework dynamics, and ecological activation mechanisms underlying the international image of Chinese EVs. By integrating unsupervised BERTopic modeling, Large Language Model (LLM) semantic mapping, the Entropy Weight Method (EWM), and Social Network Analysis (SNA), this inquiry operationalizes a comprehensive computational communication framework to mine large-scale behavioral and textual data from YouTube. The empirical findings unveil that: (1) international audience perceptions have broken through the traditional “low-cost manufacturing” stereotype, spontaneously giving rise to a multidimensional, composite cognitive schema centered on smart ecosystems and design experiences; (2) driven by the interplay of rational and irrational user feedback loops, the ecological activation efficiencies across diverse discursive dimensions exhibit pronounced nonlinear variances, characterized by a “strong activation of intelligent ecosystems versus a long-tail stagnation of cost-effectiveness salience”; and (3) positive technological frameworks and negative geopolitical or regulatory risks engage in fierce, adversarial contestation and structural hybridization within a highly volatile network topology, culminating in a unique “dual-core” configuration. Theoretically, this study enriches the scholarly understanding of country-of-origin and corporate brand images through a complex systems lens; methodologically and practically, it offers a high-fidelity, actionable quantitative paradigm for global brand empowerment and targeted cross-border public opinion governance. Full article

Background/Objectives: Artificial intelligence (AI) is increasingly used in youth mental health services, including clinical decision support, risk prediction, and digital therapeutics. However, existing frameworks provide limited guidance for integrating ethical, cultural, and relational considerations into the design, governance, and implementation of AI-enabled mental health technologies. This scoping review examined how digital humanities-informed approaches have been incorporated into AI-supported mental health interventions for children and adolescents. Methods: A scoping review was conducted following the PRISMA Extension for Scoping Reviews (PRISMA-ScR) guidelines. Peer-reviewed literature published between 2015 and 2025 was searched using PubMed and supplemented by semantic searches through Elicit. Systematic reviews, scoping reviews, and meta-analyses examining AI-enabled digital mental health interventions and digital humanities perspectives were included. Data were synthesized using inductive thematic analysis. Results: Seventeen review-level studies met the inclusion criteria. Six recurring themes were identified: engagement, participatory co-design, human oversight, equity, ethical governance, and implementation. Across the included reviews, humanities-informed approaches were associated with greater attention to relational engagement, stakeholder participation, transparency, contextual adaptation, and culturally responsive implementation. Evidence supporting intervention effectiveness was strongest in systematic reviews and meta-analyses, whereas findings related to ethics, governance, equity, and implementation were derived primarily from scoping reviews and conceptual syntheses. Conclusions: This review suggests that digital humanities provides a valuable interdisciplinary perspective for informing the design, governance, and implementation of AI-enabled youth mental health interventions. Although the current evidence base remains heterogeneous, integrating humanities-informed approaches may support the development of AI systems that are more ethical, equitable, and developmentally responsive. Future research should evaluate these approaches through empirical implementation studies and emerging generative AI applications. Full article

With the rapid development of remote sensing technology, high-resolution satellite imagery has been increasingly applied to land resource monitoring, urban planning, and environmental assessment. Automatically assigning semantic labels to remote sensing image patches remains a fundamental challenge due to pronounced intra-class variation and high inter-class visual similarity. To address the trade-off between model capacity and limited labeled data, this paper proposes a remote sensing image scene classification framework based on an improved EfficientNetV2-S architecture. The proposed model integrates a Squeeze-and-Excitation (SE) channel attention module between the final 1 × 1 expansion convolution and the Global Average Pooling layer, where it functions as a late-stage channel gating mechanism that adaptively recalibrates channel-wise responses, though its accuracy benefit is seed-sensitive rather than consistently reproducible at the current dataset scale. A two-stage optimization strategy was evaluated, comprising a fully unfrozen supervised baseline followed by a pseudo-label semi-supervised fine-tuning stage utilizing a strict confidence threshold ($\tau =0.90$). Evaluated on a 10-class subset of the public NWPU-RESISC45 benchmark, the purely supervised SE-EfficientNetV2-S delivers 98.71% independent test accuracy, matching or exceeding the much larger ResNet50 (98.50%, 24.1 M parameters) despite using only 20.4 M parameters. Multi-seed variance analysis further reveals that semi-supervised fine-tuning yields a small test-set improvement for the No-SE configuration that is consistent in sign across all three seeds (+0.46 pp mean) but not statistically significant at this sample size, and an even smaller, likewise non-significant gain for the SE-augmented model (+0.08 pp), suggesting that channel gating moderates pseudo-label effectiveness in small-data regimes. Full article

Geospatial Artificial Intelligence (GeoAI) and the rapid advancement of 3D data acquisition technologies (e.g., LiDAR) have enabled scalable semantic interpretation of georeferenced large-scale 3D point clouds across different applications. 3D deep learning-based semantic segmentation on large-scale 3D point clouds typically relies on data partitioning and sampling strategies to address computational constraints, resulting in a substantial proportion of points not being directly predicted by deep neural networks. These points not directly predicted by the model require a processing step for label propagation, which is commonly handled using simple spatial proximity-based rules. While this simplification may be acceptable for some applications, it becomes critical in tasks that require precise spatial measurements and accurate object delineation, where propagation errors can directly affect downstream analyses. To bridge this research gap, this study views this process as a spatial surrogate modeling problem, where predictions from deep learning models are used to infer labels for underrepresented points based on spatial relationships. We adopt inverse distance weighting (IDW) as a transparent, deterministic, and training-free spatial post-processing strategy to examine whether explicitly incorporating spatial relationships improves segmentation outcomes. We evaluate the proposed method within an existing 3D semantic segmentation workflow for bridge inspection, a practical application requiring accurate spatial measurement and reliable object delineation. In the experiments, we use self-collected terrestrial LiDAR point clouds of bridges and associated hydraulic structures and systematically assess how neighborhood size and distance-decay parameters affect model performance on the segmentation task. Results show that this spatial post-processing step provides a modest enhancement over the conventional nearest-neighbor propagation baseline, with notable gains especially on spatially sparse or geometrically complex classes. The results also reveal class-dependent spatial effects, suggesting that different semantic classes exhibit distinct spatial dependencies during label propagation. These findings highlight the practical importance of accounting for spatial context when propagating semantic information to those points not directly predicted by deep neural networks. This is particularly important for downstream applications that require highly accurate spatial measurement and object delineation. The practical value of this post-processing step becomes more apparent in data-scarce application domains such as hydraulic-structure inspection, where large labeled point-cloud datasets and public benchmarks remain limited. In such settings, users may rely on domain-specific pre-trained models, while the original training data may not be publicly available for retraining or extensive model modification. Using a practical case study in the hydraulic domain, this study shows that deterministic post-inference label propagation can improve complete point-wise prediction from an existing model, thereby supporting the reuse of available models for domain-specific applications. The resulting response surfaces support two practical uses: site-specific calibration when limited labeled target data are available, and empirically informed initial settings, a rule of thumb, for comparable bridge-LiDAR applications when target-site labels are unavailable. Full article

Foundation models such as the Segment Anything Model 2 (SAM2) have demonstrated strong performance in image segmentation; however, their application to industrial defect detection faces significant challenges due to the substantial domain gap between natural and industrial images, insufficient sensitivity to fine-grained high-frequency structures, and reliance on manual prompts. To address these issues, this study proposes a Hierarchical Cross-Branch Frequency-Aware Network (HCFNet) to adapt SAM2 for prompt-free industrial defect segmentation. First, a Gated Adapter is introduced into the frozen SAM2 encoder, enabling efficient cross-domain transfer without massive parameter retraining, thereby effectively preserving the pre-trained visual priors. Secondly, a Laplacian-enhanced Auxiliary Branch is designed to explicitly amplify high-frequency components, compensating for the inherent perception limitations of the Transformer backbone and significantly awakening the model’s sensitivity to subtle defects like micro-cracks. Finally, a Cross-branch Multi-scale Fusion Module is proposed to seamlessly align and integrate global semantic information with local structural details in a unified manner, resolving heterogeneous feature distribution conflicts. Extensive experiments on the MVTec AD and VisA datasets demonstrate that the proposed method consistently outperforms SAM2-based baselines in terms of mIoU and mDice. This study establishes an effective approach for leveraging foundation models in automated industrial inspection and is expected to drive advancements in precise defect perception technologies. Full article

Surface cracks pose a significant threat to the durability and operational safety of concrete bridges, making accurate crack detection essential for effective structural health monitoring. Conventional manual inspection is limited by low efficiency, subjective judgment, and safety risks during field operations. Although deep learning computer vision techniques have become the dominant approach for automated crack detection, existing review studies provide limited discussion of their adaptability to different bridge structures and offer insufficient guidance for practical engineering applications. This review systematically summarizes recent advances in deep learning methods for surface crack detection in concrete bridges. Existing approaches are classified into crack classification, object detection, and semantic segmentation. The review further examines the relationships among bridge geometry, crack morphology, inspection conditions, and model performance. It also compares the technical challenges associated with beam, arch, cable stayed, and suspension bridges and discusses suitable model adaptation strategies for different structural characteristics. The analysis shows that bridge specific model architectures can significantly improve detection accuracy and robustness under complex inspection conditions. However, several challenges remain, including the limited availability of high quality datasets, data imbalance, inadequate detection of small cracks on curved surfaces under multiple viewing conditions, and the high cost of field deployment. This review provides practical guidance for selecting appropriate deep learning techniques for concrete bridge inspection and highlights future research directions toward integrating advanced sensing, multimodal data fusion, and intelligent inspection technologies to achieve more accurate, efficient, and reliable bridge condition assessment. Full article

This article reports a structured scoping-style review, rather than a meta-analysis, of generative artificial intelligence (GenAI) applications in architectural conceptual design. Searches were conducted in Web of Science, Scopus, Dimensions and CNKI for English- and Chinese-language records, supplemented by targeted forward and backward citation chasing. The evidence base distinguishes coded studies from contextual references and comprises 58 coded studies and 15 contextual references published between 2018 and 2026. The literature was coded by technical type, design stage, input modality, output modality, evaluation strategy, editability, and stated limitation. The synthesis identifies four interrelated domains: LLM-based semantic and knowledge support, diffusion-based conceptual visualization, spatially conditioned layout and 3D scene generation, and BIM/parametric coupling within human-AI workflows. The review indicates that current research is strongest in atmospheric visualization and prompt-mediated exploration, while evidence for architectural validity, downstream editability, regulatory checking, and professional accountability remains limited. Four cross-cutting challenges—controllability, evaluability, translatability, and responsibility—are operationalized as review-derived evaluation dimensions. GenAI is therefore better understood as a representational and workflow technology for early-stage exploration than as an autonomous architectural designer. Full article

This study develops a computational framework for quantifying how technological disruptions are associated with shifts in media–public discourse alignment. Using the public release of ChatGPT (30 November 2022) as a temporal breakpoint, we examine whether the broad public availability of generative AI was associated with structural changes in the relationship between media agendas and online public discourse in South Korea. The framework combines PPMI-weighted semantic network construction with QAP correlation, MRQAP regression and Cohen’s q effect-size analysis, applied to 181,081 Korean-language texts encompassing media agendas (national newspapers, economic dailies, regional newspapers, and broadcast news) and public agendas (online communities) over six years (2019–2025). Results reveal that national newspapers lost their dominant agenda-setting position, with public alignment declining sharply (r: 0.678 → 0.430, Cohen’s q = 0.369, large effect), while economic papers rose from lowest to second-highest alignment (r: 0.352 → 0.567) by addressing market and industry dimensions that direct AI experience could not supply. Broadcasting emerged as the dominant structural anchor of public discourse (unique coefficient β: 0.263 → 0.569; its removal alone lowers model fit from R² = 0.517 to 0.347), while national newspapers’ unique contribution reversed in sign. Collective media explanatory power itself remained essentially stable (R² = 0.537 → 0.517), indicating a structural reconfiguration of which media align with public discourse rather than a wholesale weakening of media–public alignment—consistent with agenda-melding, in which publics integrate media coverage with firsthand technological experience. A residual analysis further shows that the variance media agendas leave unexplained is not noise but a structured, public-specific layer of discourse organized around the hands-on use of generative-AI tools (e.g., ChatGPT, image generation)—direct evidence of a bounded public autonomy in which public discourse is distinct from, and not reducible to, media agendas. These findings demonstrate the framework’s utility for detecting how publicly accessible AI adoption—exemplified by ChatGPT—is associated with shifts in media–public structural dynamics within discourse ecosystems and carry implications for computational social science, technology communication, and applied network analysis. Full article

Higher education plays a crucial role in equipping citizens to tackle contemporary socio-ecological challenges. However, little research has examined how different generations of university students understand socioecological transitions, or the implications of these differences for sustainability education. This study compares the perceptions of older learners (aged 55–70) enrolled in the ‘Classrooms of Experience’ programme with those of undergraduate students (aged 18–28) from the Faculty of Economics and Business at the University of the Basque Country (UPV/EHU). Qualitative data were collected from approximately 250 participants during the 2023/24 and 2024/25 academic years. The data were analysed using the Grid Elaboration Method and the IRaMuTeQ (Version 0.8 Alpha 7) text analysis tool to identify semantic structures, symbolic associations and patterns of meaning across the two age groups. The findings reveal significant generational differences in understanding socioecological transitions. Older learners tend to frame transitions as regulated processes linked to institutional action, public policy, welfare, and quality of life. In contrast, younger students interpret socioecological transitions as responses to interconnected ecological and social crises, emphasising socioecological justice, responsibility, sustainability, and technological innovation as key drivers of transformation. These results suggest the coexistence of complementary yet distinct socioecological imaginaries within the university context. The study highlights the pedagogical value of intergenerational dialogue and learning in higher education. By bringing together diverse perspectives on sustainability, universities can promote more critical, reflective, and transformative educational processes that are capable of addressing the complex challenges of contemporary socioecological transitions. Full article

Feature selection is a promising dimension reduction technology that focuses on a reduced subspace by selecting excellent features. Most existing approaches tend to emphasize the discriminative ability of features based on either a global or a local evaluation criterion alone, and a few holistic approaches explore their selection granularity beyond the instance level. This study presents a novel Semantic-aware Instance-wise Feature selection model, dubbed SIF, to address the weakness of existing methods, which assess the significance of features from an individual view. Furthermore, SIF proposes to specify feature representations at the instance level, which is rarely touched by existing methods given the considerable learning complexity. In particular, SIF is designed as a sequential pipeline framework. First, it explicitly models semantic correlations and employs this information to select semantic-aware features. Then, inconsistent instances are captured and guide the instance-wise feature selection. Both types of features constitute the final optimal feature subset, which can represent semantics at a global level as well as describe instance characteristics at a local level. An extensive experimental evaluation illustrates the superiority of SIF under various metrics. Full article

Product concept generation is essential to innovative engineering design. However, it is often constrained by designers’ domain fixation and their limited access to cross-domain technological knowledge. This paper presents PatentConceptor, an intelligent system for cross-domain innovative design. The system integrates patent knowledge retrieval with large language model (LLM) capabilities. It comprises three functional modules: design requirement analysis (DRA), patent knowledge retrieval (PKR), and product concept generation (PCG). Two principal innovations are proposed. First, a query-type-adaptive hybrid retrieval mechanism dynamically fuses TF-IDF and LSA by mapping query characteristics to optimal lexical-semantic weightings, rather than using a fixed fusion strategy. Second, a domain-constrained LLM generation paradigm enforces hard IPC-domain restrictions during concept generation, grounding concepts in retrieved patents and mitigating hallucination risk. Quantitative experiments on 50 design problems and 520,634 patent documents demonstrate the system’s strong retrieval performance (P@5 = 0.836, CP@5 = 0.212) and high concept novelty (mean novelty score = 5.29 on a 7-point scale). An end-to-end case study was conducted on a wearable system for emotional body gesture recognition. The results demonstrate that concepts generated by PatentConceptor can be successfully implemented as functional prototypes. The resulting prototype achieved a cross-subject recognition accuracy of 71.5%. Full article

Accurate detection in the early stages of a fire is a crucial prerequisite for the efficient implementation of fire suppression and emergency rescue operations. Its accuracy and timeliness directly affect the control of disaster loss severity. Traditional fire detection methods mainly include three categories, which are manual inspection, sensor detection, and visual recognition. However, manual inspection is restricted by labor costs and time efficiency, making it difficult to achieve large-scale, high-frequency and real-time fire monitoring. Sensor detection is easily interfered by environmental factors such as temperature, humidity, and dust, leading to frequent false alarms and missed alarms. Visual recognition technology has shortcomings in aspects such as detailed feature perception, dynamic scene modeling, and reasoning robustness in complex environments, making it difficult to meet the requirements of high-precision detection. To address these issues, this study innovatively proposes a lightweight fire and smoke detection model based on semantic enhancement and frequency domain perception modeling, which is named the SemaFire you only look once (SemaFire-YOLO) model. The model constructs a large language and vision assistant (LLaVA) semantic guidance module, which uses a large language model to understand and guide the semantic features of images, thereby enhancing the saliency representation intensity of small and weak target regions. Then, a Haar wavelet-based downsampling module is adopted, which compresses spatial information while preserving high-frequency features such as flame edges and smoke textures, improving the accuracy of target recognition. Next, the convolution modulation mechanism is introduced to replace the traditional attention mechanism, enhancing the overall modeling efficiency and reducing computational overhead. Finally, a Dynamic Tanh normalization module is adopted to replace the batch normalization module in the traditional YOLO algorithm, strengthening the model’s representation stability and reasoning robustness under unstable input distributions. Experimental results show that the SemaFire-YOLO model achieves a mean average precision (mAP@0.5) of 64.30% on the fire image dataset, which is 0.8, 2.0, 0.6, and 3.8 percentage points higher than that of mainstream models such as YOLOv5n, YOLOv8n, YOLOv11n, and YOLOv12n, respectively. It exhibits better boundary detection capability and practical deployment potential. Through visual analysis, the results indicate that the improved SemaFire-YOLO model achieves more accurate detection and higher confidence in actual complex scenarios, further verifying the model’s robustness and accuracy in complex scenarios such as low contrast and dynamic fire conditions. Full article

Trajectory Representations Learning (TRL) serves as a foundational technology for supporting intelligent transportation. However, models trained on real-world data often suffer from performance degradation caused by inherent spatiotemporal distribution bias, which reflects the heterogeneity of urban structures and human movement behaviors. This leads to representations that overfit to frequent patterns, resulting in weak robustness and limited generalization to sparse or atypical trajectories. To address these issues, this paper presents a novel perspective, anchor-guided balanced learning, and instantiates it with a framework, AnchorTRL. AnchorTRL introduces anchors to proactively construct a balanced semantic space instead of passively fitting the empirical data distribution. Specifically, AnchorTRL designs a spatiotemporal anchor identification algorithm to recognize trajectory anchors that comprehensively cover the data manifold. And, it proposes a calculation method to measure all trajectories’ semantic similarity with anchors. Additionally, it develops an anchor-based balanced sampling strategy to mitigate the dominance of frequent patterns and steer the model towards learning a more balanced representation. Finally, it constructs a multi-task contrastive learning objective with adaptive constraints to enhance the aggregation of semantically similar trajectories. Experimental results show that AnchorTRL outperforms existing baseline methods in tasks such as travel time estimation and similar trajectory queries, demonstrating its effectiveness and robustness. This research provides methodological support for constructing more reliable trajectory representation learning models, and offers new insights for optimizing intelligent transportation applications under spatiotemporal biases. Full article

This article develops a conceptual account of semantic displacement in conversational AI. The central question concerns how agency is affected when systems do more than automate information retrieval and begin to supply the descriptions, classifications and normative cues through which users understand what they are doing. Drawing on philosophy of action, philosophy of language, hermeneutics, philosophy of technology and critical accounts of algorithmic mediation, this article reconstructs the relation between meaning and action as a condition of agency. Its methodological approach is conceptual and diagnostic, oriented toward clarifying a problem that becomes visible when established theories are brought together in relation to contemporary conversational systems. The article interprets these systems as operational semantic infrastructures that organize context-sensitive linguistic uptake within practical environments such as health, work, education, administration and everyday self-management. It then introduces semantic displacement as the condition in which action-relevant meanings become increasingly organized, prioritized and consolidated outside the agent’s own participatory interpretation. The argument contributes a vocabulary for distinguishing agency-enhancing semantic support from forms of semantic substitution that weaken interpretive participation. It concludes by proposing semantic sovereignty and interpretive contestability as normative ideals for human agency in AI-mediated environments. The argument specifies action as intentional conduct understood under socially available descriptions and cognition as situated interpretive sense-making rather than purely internal computation. It also clarifies three conditions under which semantic support becomes displacement: opaque semantic generation, practical stabilization and reduced interpretive contestability. Full article

Distributed energy systems are increasingly important for strengthening critical infrastructure resilience under conditions of technological, climatic, geopolitical, and cyber disruption. However, existing research on energy resilience is still dominated by technical approaches focused on reliability, renewable energy integration, microgrid control, and storage optimisation, while the role of distributed energy systems in the full crisis-management cycle remains insufficiently conceptualised. This article addresses this gap by combining a scoping review, lexicographic and semantic analysis using IRaMuTeQ version 0.7 alpha 2, and a conceptual-methodological framework for assessing distributed energy systems as instruments of crisis management. The main contribution of the study is the M_ZK-DES model, which integrates technological-infrastructural, decision-operational, legal-institutional, and socio-organisational dimensions with four crisis-management phases: prevention, preparedness, response, and recovery. The model distinguishes distributed energy systems, distributed energy resources, distributed generation, microgrids, prosumers, energy communities, and energy clusters and links them to measurable resilience indicators. These include SAIDI, SAIFI, energy not supplied, restoration time, share of critical load served, islanding capability, voltage and frequency stability, storage autonomy, procedural readiness, and local coordination capacity. The analysis shows that distributed energy systems may reduce vulnerability to cascading failures, support islanded operation, protect vulnerable consumers, improve emergency power continuity, and strengthen local energy autonomy. The proposed scoring and weighting logic enables future empirical validation, scenario testing, and comparative assessment across regions and crisis types, including extreme weather events, cyberattacks, and supply-chain disruptions. The article contributes to energy resilience and crisis-management studies by offering an integrated and operational framework for evaluating distributed energy systems as practical tools for critical infrastructure protection and continuity of essential public services. Full article