Search Results (14,436)

In order to increase the effectiveness of model training, data reduction is essential to data-centric Artificial Intelligence (AI). It achieves this by locating the most instructive examples in massive datasets. To increase data quality and training efficiency, the main difficulty is choosing the best examples rather than the complete datasets. In this paper, we propose an effective data reduction strategy based on Pointwise 𝒱-Information (PVI). To enable a static method, we first use PVI to quantify instance difficulty and remove instances with low difficulty. Experiments show that classifier performance is maintained with only a 0.0001% to 0.76% decline in accuracy when 10–30% of the data is removed. Second, we train the classifiers using a progressive learning strategy on examples sorted by increasing PVI, accelerating convergence and achieving a 0.8% accuracy gain over conventional training. Our findings imply that training a classifier on the chosen optimal subset may improve model performance and increase training efficiency when combined with an efficient data reduction strategy. Furthermore, we have adapted the PVI framework, which was previously limited to English datasets, to a variety of Chinese Natural Language Processing (NLP) tasks and base models, yielding insightful results for faster training and cross-lingual data reduction. Full article

Vegetation dynamics are complexly influenced by multiple factors such as climate, human activities, and topography. In recent years, the frequency, intensity, and diversity of human activities have increased, placing substantial pressure on the growth of vegetation. Arid and semi-arid regions are particularly sensitive to climate change, and climate change and large-scale ecological restoration have led to significant changes in the dynamic of dryland vegetation. However, few studies have explored the nonlinear relationships between these factors and vegetation dynamic. In this study, we integrated trend analysis (using the Mann–Kendall test and Theil–Sen estimation) and machine learning algorithms (XGBoost-SHAP model) based on long time-series remote sensing data from 2001 to 2020 to quantify the nonlinear response patterns and threshold effects of bioclimatic variables, topographic features, soil attributes, and anthropogenic factors on vegetation dynamic. The results revealed the following key findings: (1) The kNDVI in the study area showed an overall significant increasing trend (p < 0.01) during the observation period, of which 26.7% of the area showed a significant increase. (2) The water content index (Bio 23, 19.6%), the change in land use (15.2%), multi-year average precipitation (pre, 15.0%), population density (13.2%), and rainfall seasonality (Bio 15, 10.9%) were the key factors driving the dynamic change of vegetation, with the combined contribution of natural factors amounting to 64.3%. (3) Among the topographic factors, altitude had a more significant effect on vegetation dynamics, with higher altitude regions less likely to experience vegetation greening. Both natural and anthropogenic factors exhibited nonlinear responses and interactive effects, contributing to the observed dynamic trends. This study provides valuable insights into the driving mechanisms behind the condition of vegetation in arid and semi-arid regions of China and, by extension, in other arid regions globally. Full article

In response to the stress commonly experienced by young people in high-pressure daily environments, a music-based stress-relief interactive system was developed by integrating music-assisted care with emotion-sensing technology. The design principles of the system were established through a literature review on stress, music listening, emotion detection, and interactive devices. A prototype was created accordingly and refined through interviews with four experts and eleven users participating in a preliminary experiment. The system is grounded in a four-stage guided imagery and music framework, along with a static activity model focused on relaxation-based stress management. Emotion detection was achieved using a wearable EEG device (NeuroSky’s MindWave Mobile device) and a two-dimensional emotion model, and the emotional states were translated into visual representations using seasonal and weather metaphors. A formal experiment involving 52 users was conducted. The system was evaluated, and its effectiveness confirmed, through user interviews and questionnaire surveys, with statistical analysis conducted using SPSS 26 and AMOS 23. The findings reveal that: (1) integrating emotion sensing with music listening creates a novel and engaging interactive experience; (2) emotional states can be effectively visualized using nature-inspired metaphors, enhancing user immersion and understanding; and (3) the combination of music listening, guided imagery, and real-time emotional feedback successfully promotes emotional relaxation and increases self-awareness. Full article

In this work, the utility of multimodal vision–language models (VLMs) for visual product understanding in e-commerce is investigated, focusing on two complementary models: ColQwen2 (vidore/colqwen2-v1.0) and ColPali (vidore/colpali-v1.2-hf). These models are integrated into two architectures and evaluated across various product interpretation tasks, including image-grounded question answering, brand recognition and visual retrieval based on natural language prompts. ColQwen2, built on the Qwen2-VL backbone with LoRA-based adapter hot-swapping, demonstrates strong performance, allowing end-to-end image querying and text response synthesis. It excels at identifying attributes such as brand, color or usage based solely on product images and responds fluently to user questions. In contrast, ColPali, which utilizes the PaliGemma backbone, is optimized for explainability. It delivers detailed visual-token alignment maps that reveal how specific regions of an image contribute to retrieval decisions, offering transparency ideal for diagnostics or educational applications. Through comparative experiments using footwear imagery, it is demonstrated that ColQwen2 is highly effective in generating accurate responses to product-related questions, while ColPali provides fine-grained visual explanations that reinforce trust and model accountability. Full article

Hydrogen sulfide (H₂S) can be produced from the metabolism of foods containing sulfur in the gastrointestinal tract (GIT). At low doses, H₂S regulates the gut microbial community and supports GIT health, but depending on dose, age, and individual health conditions, it may also contribute to inflammatory responses and gut barrier dysfunction. Controlling H₂S production in the GIT is important for maintaining a healthy gut microbiome. However, research on this subject is limited due to the gaseous nature of the chemical and the difficulty of accessing the GIT in situ. In the present ex vivo experiment, we used a single-dose sodium sulfide preparation (SSP) as a H₂S precursor to test the effect of H₂S on the human gut microbiome across different age groups, including breastfed infants, toddlers, adults, and older adults. Metagenomic sequencing and metabolite measurements revealed that the development of the gut microbial community and the production of short-chain fatty-acids (SCFAs) were age-dependent; that the infant and the older adult groups were more sensitive to SSP exposure; that exogeneous SSP suppressed SCFA production across all age groups, except for butyrate in the older adult group, suggesting that H₂S selectively favors specific gut microbial processes. Full article

Promoting sustainable development requires a clear understanding of how short-term fluctuations in anthropogenic emissions affect urban environmental quality. This is especially relevant for cities experiencing rapid industrial changes or emergency policy interventions. Among key environmental concerns, variations in ambient pollutants like ozone (O₃) are closely tied to both public health and long-term sustainability goals. However, traditional chemical transport models often face challenges in accurately estimating emission changes and providing timely assessments. In contrast, statistical approaches such as the difference-in-differences (DID) model utilize observational data to improve evaluation accuracy and efficiency. This study leverages the synthetic difference-in-differences (SDID) approach, which integrates the strengths of both DID and the synthetic control method (SCM), to provide a more reliable and accurate analysis of the impacts of interventions on city-level air quality. Using Shanghai’s 2022 lockdown as a case study, we compare the deweathered ozone (O₃) concentration in Shanghai to a counterfactual constructed from a weighted average of cities in the Yangtze River Delta (YRD) that did not undergo lockdown. The quasi-natural experiment reveals an average increase of 4.4 μg/m³ (95% CI: 0.24–8.56) in Shanghai’s maximum daily 8 h O₃ concentration attributable to the lockdown. The SDID method reduces reliance on the parallel trends assumption and improves the estimate stability through unit- and time-specific weights. Multiple robustness checks confirm the reliability of these findings, underscoring the efficacy of the SDID approach in quantitatively evaluating the causal impact of emission perturbations on air quality. This study provides credible causal evidence of the environmental impact of short-term policy interventions, highlighting the utility of SDID in informing adaptive air quality management. The findings support the development of timely, evidence-based strategies for sustainable urban governance and environmental policy design. Full article

Artificial Light at Night (ALAN) is a recent issue of concern for researchers primarily working on the anthropogenic impacts on animal and ecosystem health. Our concern is associated with the ALAN exposure to an aquatic ecosystem by disrupting the natural dark–light cycle, which is essential for maintaining the overall health of the ecosystem and its inhabitants. In this study, we have attempted to understand the adverse consequences of ALAN in inducing neuro-behavioural stress in a freshwater prawn species (aquatic arthropod) Macrobrachium lamarrei by considering grooming behaviour, a well-established indicator of neurological stress in animals. Our results show that continuous ALAN exposure (for seven days) can increase collective grooming activity in Macrobrachium lamarrei over time. In our experiment, we have used two intensities of ALAN (50 and 120 lux). Although the response (in terms collective grooming) to both intensities are apparently different, our fundamental hypothesis is confirmed, where it is evident that prolonged light exposure can induce an elevation in cumulative grooming performances in a freshwater prawn population. Full article

Amid intensifying global ESG regulations and the expanding influence of green finance, China’s digital economy policies have emerged as key institutional instruments for promoting corporate sustainability. Leveraging the implementation of the National Big Data Comprehensive Pilot Zone as a quasi-natural experiment, this study utilizes panel data of Chinese listed firms from 2009 to 2023 and applies multi-period Difference-in-Differences (DID) and Spatial DID models to rigorously identify the policy’s effects on corporate ESG performance. Empirical results indicate that the impact of digital economy policy is not exerted through a direct linear pathway but operates via three institutional mechanisms, enhanced information transparency, eased financing constraints, and expanded fiscal support, collectively constructing a logic of “institutional embedding–governance restructuring.” Moreover, disruptive technological innovation significantly amplifies the effects of the transparency and fiscal mechanisms, but exhibits no statistically significant moderating effect on the financing constraint pathway, suggesting a misalignment between innovation heterogeneity and financial responsiveness. Further heterogeneity analysis confirms that the policy effect is concentrated among firms characterized by robust governance structures, high levels of property rights marketization, and greater digital maturity. This study contributes to the literature by developing an integrated moderated mediation framework rooted in institutional theory, agency theory, and dynamic capabilities theory. The findings advance the theoretical understanding of ESG policy transmission by unpacking the micro-foundations of institutional response under digital policy regimes, while offering actionable insights into the strategic alignment of digital transformation and sustainability-oriented governance. Full article

The justice system has increasingly applied AI techniques for legal judgment to enhance efficiency. However, most AI techniques focus on decision-making outcomes, failing to capture the deliberative nature of the real-world judicial process. To address these challenges, we propose a large language model-based multi-agent framework named AgentsBench. Our approach leverages multiple LLM-driven agents that simulate the discussion process of the Chinese judicial bench, which is often composed of professional and lay judge agents. We conducted experiments on a legal judgment prediction task, and the results show that our framework outperforms existing LLM-based methods in terms of performance and decision quality. By incorporating these elements, our framework reflects real-world judicial processes more closely, enhancing accuracy, fairness, and societal consideration. While the simulation is based on China’s lay judge system, our framework is generalizable and can be adapted to various legal scenarios and other legal systems involving collective decision-making processes. Full article

Red blood cell (RBC) production from bone marrow hematopoietic stem cells (BMHSCs) in vitro overlooks the mechanical signals of the bone marrow niche and overly relies on growth factors. Considering that the fate of hematopoietic stem cells (HSCs) is determined by the natural bone marrow microenvironment, differences in mechanical microenvironments provide a reference for the regulation of HSC differentiation. This study seek to reveal the role of mechanobiology cues in erythropoiesis and provide a new perspective for the design of in vitro erythropoiesis platforms. The hydrogel platforms we designed simulate the stiffness gradient of the bone marrow niche to culture HSCs and induce their differentiation into the erythroid system. Cells on the low-stiffness scaffold have higher potential for erythrocyte differentiation and faster differentiation efficiency and promote erythrocyte differentiation after erythropoietin (EPO) restriction. In vivo transplantation experiments demonstrated that these cells have the ability for continuous proliferation and differentiation into mature erythrocytes. By combining mechanical cues with in vitro erythrocyte production, this method is expected to provide insights for in vitro hematopoietic design and offer a scalable cell manufacturing platform for transfusion medicine. Full article

As a key engine driving China’s green financial transformation, the Green Financial Reform and Innovation Pilot Zones have demonstrated significant achievements in enhancing the capacity of financial services to support green real economies, preventing and mitigating green financial risks, and bolstering national and urban economic resilience. On this basis, a spatial Markov chain model is applied to further analyze the economic toughness of prefecture-level cities. This study treats the establishment of these pilot zones as a quasi-natural experiment, using panel data from 269 prefecture-level cities in China from 2013 to 2023 and employing a multi-period difference-in-differences (DID) model to empirically examine the impact of green financial reform on urban economic resilience and its underlying mechanisms. The results reveal that the establishment of these pilot zones significantly enhances urban economic resilience. Specifically, green financial reforms primarily improve urban economic resilience by increasing credit accessibility and capital allocation efficiency in the pilot cities. Furthermore, the policy effects are more pronounced in large cities and resource-dependent cities compared to small and medium-sized cities and non-resource-dependent cities, with stronger impacts observed in southern and coastal regions than in northern inland areas. Additionally, the policy effects are significantly greater in environmentally prioritized cities than in non-prioritized cities. By integrating green financial reforms and urban economic resilience into a unified analytical framework, this study provides valuable insights for policymakers to refine green financial strategies and design resilience-enhancing policies. Full article

During the course of genome mining initiatives, we identified a marine-derived Micromonospora, assigned here as strain WMMD956; the genome of WMMD956 appeared to contain a number of features associated with everninomicins, well-known antimicrobial orthosomycins. In addition, LCMS-based hierarchical clustering analysis and principal component analysis (hcapca) revealed that WMMD956 displayed an extreme degree of metabolomic and genomic novelty. Dereplication of high-resolution tandem mass spectrometry (HRMS/MS) and Global Natural Product Social molecular networking platform (GNPS) analysis of WMMD956 resulted in the identification of several analogs of the previously known everninomicin. Chemical structures were unambiguously confirmed by HR-ESI-MS, 1D and 2D NMR experiments, and the use of MS/MS data. The isolated metabolites, 1–3, were evaluated for their antibacterial activity against methicillin-resistant Staphalococcus aureus (MRSA). Full article

High-quality green gardens can markedly enhance the quality of life and mental well-being of their users. However, health and lifestyle constraints make it difficult for people to enjoy urban gardens, and traditional methods struggle to offer the high-fidelity experiences they need. This study introduces a 3D scene reconstruction and rendering strategy based on implicit neural representation through the efficient and removable neural radiation fields model (NeRF-RE). Leveraging neural radiance fields (NeRF), the model incorporates a multi-resolution hash grid and proposal network to improve training efficiency and modeling accuracy, while integrating a segment-anything model to safeguard public privacy. Take the crabapple tree, extensively utilized in urban garden design across temperate regions of the Northern Hemisphere. A dataset comprising 660 images of crabapple trees exhibiting three distinct geometric forms is collected to assess the NeRF-RE model’s performance. The results demonstrated that the ‘harvest gold’ crabapple scene had the highest reconstruction accuracy, with PSNR, LPIPS and SSIM of 24.80 dB, 0.34 and 0.74, respectively. Compared to the Mip-NeRF 360 model, the NeRF-RE model not only showed an up to 21-fold increase in training efficiency for three types of crabapple trees, but also exhibited a less pronounced impact of dataset size on reconstruction accuracy. This study reconstructs real scenes with high fidelity using virtual reality technology. It not only facilitates people’s personal enjoyment of the beauty of natural gardens at home, but also makes certain contributions to the publicity and promotion of urban landscapes. Full article

The exponential increase in textual data generated across sectors such as healthcare, finance, and smart manufacturing has intensified the need for effective Big Data analytics. Large language models (LLMs) have become critical tools because of their advanced language processing capabilities. However, their static nature limits their ability to incorporate real-time and domain-specific knowledge. Retrieval-augmented generation (RAG) addresses these limitations by enriching LLM outputs through external content retrieval. Nevertheless, traditional RAG systems remain inefficient, often exhibiting high retrieval latency, redundancy, and diminished response quality when scaled to large datasets. This paper proposes an innovative structured RAG framework specifically designed for large-scale Big Data analytics. The framework transforms unstructured partial prompts into structured semantically coherent partial prompts, leveraging element-specific embedding models and dimensionality reduction techniques, such as principal component analysis. To further improve the retrieval accuracy and computational efficiency, we introduce a multi-level filtering approach integrating semantic constraints and redundancy elimination. In the experiments, the proposed method was compared with structured-format RAG. After generating prompts utilizing two methods, silhouette scores were computed to assess the quality of embedding clusters. The proposed method outperformed the baseline by improving the clustering quality by 32.3%. These results demonstrate the effectiveness of the framework in enhancing LLMs for accurate, diverse, and efficient decision-making in complex Big Data environments. Full article

As the fundamental physical carrier for human production and socio-economic endeavors, enhancing urban land green use efficiency (ULGUE) is crucial for realizing sustainable development. To effectively enhance urban land green use efficiency, this study systematically examines the intrinsic relationship between industrial policies and ULGUE based on panel data from 286 Chinese cities (2010–2022), employing an integrated methodology that combines the Difference-in-Differences (DID) model, Super-Efficiency Slacks-Based Measure Data Envelopment Analysis model, and ArcGIS spatial analysis techniques. The findings clearly demonstrate that the establishment of the “Made in China 2025” pilot policy significantly improves urban land green use efficiency in pilot cities, a conclusion that endures following a succession of stringent evaluations. Moreover, studying its mechanisms suggests that the pilot policy primarily enhances urban land green use efficiency by promoting industrial upgrading, accelerating technological innovation, and strengthening environmental regulations. Heterogeneity analysis further indicates that the policy effects are more significant in urban areas characterized by high manufacturing agglomeration, non-provincial capital/non-municipal status, high industrial intelligence levels, and less sophisticated industrial structure. This research not only provides valuable policy insights for China to enhance urban land green use efficiency and promote high-quality regional sustainable development but also offers meaningful references for global efforts toward advancing urban sustainability. Full article