Search Results (24,509)

Enhancing building energy performance has become an essential goal, particularly as building energy management systems (BEMSs) increasingly rely on high-quality data and reliable predictive models. Although machine learning (ML) models have been widely applied to building energy prediction, optimisation, and management, their reliability in practice is often constrained by data preprocessing rather than algorithm selection. Existing studies often emphasise algorithmic development while providing limited systematic investigation of preprocessing practices, leading to methodological misconceptions and reduced robustness of ML-driven building energy management. As a novel contribution, this article presents a systematic review of 73 scientific articles published from 2020 to 2025 in the field of preprocessing practices. To this goal, a three-step data preprocessing workflow is organised, comprising data analysis, data preparation, and feature engineering. The strengths, limitations, and recurring misconceptions of preprocessing techniques adopted in the analysed studies are synthesised, with emphasis on their impact on prediction accuracy, interpretability, and model robustness. As a result, this review reframes the data preprocessing stage as a decision-making process in which data analysis and the energy improvement task constrain and inform subsequent data preparation and feature engineering steps to address building energy performance enhancement tasks. Full article

Real-time measurement of tea bud phenotypes via mobile devices is constrained by model lightweighting challenges, and research on non-contact measurement of tea bud phenotypes based on key points remains largely unexplored. Information on the growth posture of tea buds is an important basis for determining tea maturity grades, quality monitoring, and tea breeding. Therefore, this work develops a deep learning-enabled YOLOv8p-Tea model to estimate key point information of tea bud posture and automatically obtain three-dimensional point cloud information of tea buds by integrating depth information, thereby achieving in situ measurement of tea bud phenotypic parameters. Meanwhile, the model is trained and validated using a tea bud (one-bud-three-leaf) image dataset, and its effectiveness is demonstrated through experiments. Compared to the YOLOv8p-pose model, the model achieves a mAP50 of 98.3%, a P of 97%, and parameters of 0.72 M, with mAP50 and P improved by 1.5% and 1.9%, respectively, and the parameter count is reduced by 25%. To validate the accuracy of phenotypic extraction, the model was deployed on edge devices, and 30 tea buds with one bud and three leaves were randomly selected in a tea garden. The final in situ measurement results showed an MRE of 6.63%. Experimental findings indicate that the developed method is capable of not only effectively estimate tea bud posture but also accurately achieves in situ measurement of tea bud phenotypes, which holds potential applications for meeting the construction needs of smart tea gardens and optimizing tea breeding. Full article

Produced-water (PW) management in the Permian Basin faces tightening injection constraints, induced seismicity concerns, and volatile saltwater disposal (SWD) costs. At the same time, chemistry-rich PW contains dissolved constituents (e.g., Li, B, and Sr) that may be valorized if SWD recovery performance and market conditions support favorable techno-economics. Here, we develop an integrated decision-support framework that couples (i) chemistry-informed surrogate models for unit process performance (recovery, effluent quality, and energy/chemical intensity) with (ii) a network-based allocation model that routes PW from sources through pretreatment, optional treatment and mineral-recovery modules (e.g., desalination and direct lithium extraction), and end-use nodes (beneficial reuse, hydraulic fracturing reuse, mineral recovery/valorization, or Class II disposal). This is a screening-level demonstration using publicly available chemistry percentiles and representative pilot-reported performance windows; it is not a site-specific facility design or a bankable TEA for a particular operator. The optimization is posed as a tri-objective problem—to maximize expected net present value, minimize SWD, and minimize an injection-risk indicator R—subject to mass balance, capacity, quality, and regulatory constraints. Uncertainty in commodity prices, recovery fractions, and operating costs is propagated via Monte Carlo scenario sampling, yielding PARETO-efficient portfolios that quantify trade-offs between profitability and risk mitigation. Using the PW chemistry percentiles reported by the Texas Produced Water Consortium for the Delaware and Midland Basins, we derive screening-level break-even lithium concentrations and illustrate how lithium-carbonate-equivalent price and recovery govern the extent to which mineral revenue can offset SWD expenditures. Comparative brine benchmarks (Smackover Formation and Salton Sea geothermal systems) contextualize the Permian’s generally lower-Li PW and highlight transferability of the workflow across brine types. The proposed framework provides a transparent, extensible basis for design matrix planning under evolving injection limits, enabling risk-aware PW management strategies that reduce disposal dependence while improving water resilience. Full article

Optical attenuation caused by absorption and scattering in turbid water significantly degrades underwater image quality, making reliable underwater imaging a challenging problem. Underwater polarization imaging has attracted increasing attention because of its ability to suppress scattered light and provide additional polarization cues. However, existing polarization-based enhancement approaches often adapt conventional underwater image enhancement strategies, and the multi-dimensional characteristics of polarization information are not always fully utilized, which may limit detail restoration in complex underwater environments. To address this issue, this paper proposes a bio-inspired underwater polarization image enhancement framework motivated by the polarization vision mechanism of marine organisms. Specifically, a two-stage architecture consisting of a Polarization Adversarial Network (PAN) and a Polarization Enhancement Network (PEN) is designed. The PAN incorporates a Bionic Antagonistic Module (BAM) to exploit complementary information among polarization channels, while Salient Feature Extraction (SFE) is introduced to reduce redundant feature interference. The subsequent PEN integrates a frequency-aware Mamba-based structure to enhance feature representation and improve detail reconstruction. Experiments on simulated underwater polarization datasets indicate that the proposed framework can effectively suppress backscattering and improve structural detail visibility in challenging underwater scenes, demonstrating competitive performance compared with representative traditional and learning-based methods. Full article

Sewerage systems are among the most fundamental and indispensable components of urban infrastructure. However, inadequate management can result in malfunctions and subsequent rehabilitation processes, leading to various negative consequences. Identifying areas at high risk of failure and conducting system-based inspections can significantly improve the performance of sewer networks. This study identified and categorized 33 criteria that could cause sewer system failures: structural, operational, hydraulic and environmental defects. A Bayesian network (BN) model was developed to determine dependencies between the criteria, quantify uncertainty, investigate new information about the structural condition of assets and calculate the effects and sensitivities of the criteria on the probability of failure. A probability-based risk assessment model was then created using a fuzzy inference system (FIS) to predict risk levels in sewerage systems under different combinations of physical and operational conditions and hydraulic and environmental effects. A case study was performed on a sewer network in Malatya, Turkey, determining its failure probability to be 76.6%, placing it in the high-risk category. When the probability of pipe failure was set to 100% in the Bayesian network model to evaluate the relative influence of different criteria, the most influential factors were identified as flow velocity (74.8%), clogging (71.4%), and failure rate (71.1%). Thanks to the flexible structure of BNs, the proposed model is expected to be useful for performing risk analyses in systems involving uncertainty or missing data. It can also be used to prioritize rehabilitation, inspection and maintenance programs, improve infrastructure service quality and ensure system reliability in urban sewerage systems. Full article

Adverse weather and low illumination remain major challenges for autonomous driving perception, where semantic segmentation must stay reliable despite severe appearance degradation. In unsupervised domain adaptation without target annotations, self-training is widely used, but it is often limited by the inconsistent quality of teacher-generated pseudo labels across samples, regions, and training stages. This paper presents RaDA, a reliability-aware self-training framework that regulates pseudo supervision at three levels. First, a progressive exposure strategy determines which target images are admitted for training. Second, spatial reliability weighting suppresses gradients from degraded regions while retaining informative supervision. Third, adaptive teacher update scheduling stabilizes pseudo label generation over time. Experiments on real-world adverse driving benchmarks show that RaDA improves robustness, training stability, and cross-dataset generalization compared with strong baselines. Compared with the previous state-of-the-art method MIC, RaDA achieves mIoU gains of 10.6 percentage points on Foggy Zurich and 8.8 percentage points on the Foggy Driving benchmark. These results indicate that explicit reliability regulation can strengthen self-training domain adaptation for semantic segmentation in autonomous driving under challenging environmental conditions. Full article

The emergence of Generative Artificial Intelligence (GenAI), capable of interpreting and reasoning with human language, has catalyzed a paradigm shift across various societal sectors. Within the tourism industry, GenAI is increasingly utilized to facilitate personalized itinerary planning, destination recommendations, and the provision of optimal route information. This study evaluates the reliability of GenAI in identifying the nearest metro station within a walking distance from tourist attractions in Busan, South Korea. Furthermore, it aims to empirically verify the determinants influencing the correctness of AI-generated responses compared to network-based shortest-path analyses. The empirical results demonstrate that Google’s Gemini 3 Pro model achieved superior performance, recording an accuracy rate of 65.0%. Regression analysis revealed that for both Gemini and GPT models, the volume of news articles associated with an attraction—representing media visibility—significantly increased the likelihood of accurate information provision. Notably, the Gemini model exhibited distinct sensitivity to geographic factors and text similarity metrics, suggesting a difference in how it processes spatial context compared to other models. Consequently, this study underscores the importance of high-quality AI-generated tourism data and offers significant contributions to the advancement of sophisticated personalized travel planning systems and GeoAI research focused on spatial problem-solving. Full article

Engagement is essential in informal learning contexts, as it fosters meaningful learning, personal relevance, and sustained motivation. However, engagement is a complex construct that requires diverse methodological approaches for accurate assessment. This study empirically examines a multi-method approach, combining questionnaires, log file analyses, and observations, in the context of a tablet-based educational program developed for an exhibition on mobility and traffic. A total of 307 tenth-grade students from 21 classes at twelve state high schools participated in self-regulated learning activities during the museum visit. Findings reveal that each method offers distinct insights: questionnaires capture students’ self-reported engagement, log files track behavioral patterns, and observations provide qualitative evidence of interaction quality. Each method provides valuable, specific insights into student engagement. Thus, integrating multiple approaches yields a more comprehensive understanding of engagement. These results underscore the importance of methodological pluralism and critical reflection when interpreting research findings from different methodological sources in informal learning environments. Full article

Environmental, social, and governance (ESG) information is increasingly vital in driving capital markets to promote sustainable development. However, significant barriers remain in effectively transforming ESG information into intelligence that supports investor decision-making. Drawing upon information chain theory and intelligence transformation theory, this study constructs an ESG intelligence transformation model tailored for investor decision-making, aiming to address relevant challenges within China’s unique capital market environment. Through a mixed deductive-inductive content analysis of analyst reports issued by Chinese securities firms, this study identifies underlying issues in current ESG information utilization: excessive focus on social dimensions at the expense of integrated consideration of environmental and governance issues; inadequate conversion of environmental and governance data into decision-relevant information; and incomplete pathways for transforming ESG knowledge into intelligence supporting investment decisions. These constraints significantly undermine the potential of ESG information to guide sustainable investment strategies and support green economic transformation. To bridge these gaps, this study proposes an integrated multi-stakeholder optimization strategy encompassing: enhanced disclosure standards by regulators; greater corporate emphasis on environmental and governance disclosures; more sophisticated assessment techniques by rating agencies; and optimized information channels between corporations and investors by financial institutions. This study provides a theoretical foundation and practical pathways for enhancing the quality and utility of ESG information, contributing to sustainable finance research. Full article

In complex ground environments, conventional RRT* often suffers from poor path quality and slow expansion during robot path planning. To address these issues, this paper proposes GEAR-RRT* (Goal-guided, adaptive informed-Ellipse sampling, layered obstacle-Avoidance expansion, and cost-driven Rewiring), which constructs a collaborative optimization mechanism across the three stages of sampling, expansion, and rewiring. First, the proposed method employs an adaptive informed ellipse to concentrate sampling within feasible regions while dynamically adjusting the informed-ellipse sampling domain, and further integrates Halton-directional hybrid sampling to generate high-quality candidate samples within that domain. Meanwhile, a layered expansion strategy is adopted: the planner first performs direct goal connection for rapid progress toward the goal; when this expansion is blocked by obstacles, it switches to local multi-directional offset to search for feasible expansion directions; if this still fails, an adaptive Artificial Potential Field is introduced to guide subsequent expansions until a feasible path is found. Next, a multi-factor rewiring parent selection strategy is used to optimize path length, safety clearance, and turning angle, while cubic B-spline smoothing is applied to improve path continuity. Finally, GEAR-RRT* is evaluated in five simulation environments as well as in joint ROS and physical-robot validation and is compared with five improved RRT* variants. The results demonstrate that the proposed method achieves superior overall performance in planning time, path length, and safety clearance. Full article

Background: Dentists constitute one of the most heavily burdened groups of healthcare professionals, experiencing high levels of musculoskeletal disorders, occupational stress, burnout, and diminished quality of life. Although extensive literature addresses these issues, no study has directly examined biological age or epigenetic markers of aging in this population. This narrative review, informed by systematic methodological principles, seeks to fill this gap by connecting established occupational stressors with contemporary concepts of biological aging and chronomedicine, ultimately proposing a preventive well-being framework specifically for dentists. Methods: A narrative review informed by systematic methodology was conducted following PRISMA 2020 guidelines. Searches in PubMed, Scopus, and the Cochrane Library (2015–2025) used combined keywords and MeSH terms related to lifestyle factors, occupational stress, musculoskeletal disorders, quality of life, and wellness among dentists. Of the 943 records identified, 15 met the inclusion criteria and were assessed for outcomes, methodological quality, and relevant risk factors. Results: The included studies consistently indicated a significant occupational burden, with musculoskeletal pain, emotional exhaustion, anxiety, and depersonalization as frequent findings. Quality of life was generally moderate to low, especially regarding mental health. Lifestyle patterns were characterized by inadequate sleep, limited physical activity, irregular eating habits, and insufficient recovery. These conditions-chronic stress, poor sleep, inactivity, and suboptimal nutrition-are recognized accelerators of biological aging, implying that the professional demands of dentistry may adversely influence the biological clock. Although none of the studies measured biological age directly, the collective evidence underscores the need for preventive strategies informed by chronomedicine. Conclusions: This review highlights a critical gap in the dental literature: the complete absence of biological-age assessment in a professional population exposed to multiple aging accelerators. Integrating occupational health data with modern concepts of biological aging and chronomedicine, the study proposes a targeted preventive framework to regulate biological rhythms, reduce cumulative biological deterioration, and improve the long-term quality of life and professional sustainability of dentists. Full article

In the field of digital architectural heritage, the mining of tacit local knowledge embedded in architectural heritage is considered essential for the preservation, inheritance, and application of regional architectural characteristics. Local knowledge can be formally represented through semantic models, by which the automated mining of tacit information can be facilitated. However, due to the incomplete preservation of physical buildings and the fragmented nature of historical records, local knowledge is often represented as semantic fragments. Consequently, existing semantic models are still challenged in terms of knowledge integration and reasoning. In this study, a knowledge graph was developed for representing local knowledge, in which fragmented local semantics were aligned at both the ontological and entity levels. Subsequently, implicit local knowledge mining is achieved through meta-path centrality propagation combined with expert evaluation on a graph visualization platform. The method was applied to eight historical buildings in a case study. The knowledge graph quality assessment results indicate excellent ontology utilization and property utilization. The knowledge mining results demonstrate that graph-based expert evaluation successfully enables knowledge Feature Ranking and knowledge Extinction Warning. Full article

Doctoral training in applied computing and information systems is the structured development of a researcher’s capacity to produce original, rigorous, and scholarship that is relevant to practice, supported through doctoral supervision, which provides academic guidance for research design decisions, progress management, scholarly quality, and researcher development. In this setting, Design Science Research (DSR) is a methodology that generates knowledge through the purposeful design and evaluation of an artifact intended to address a defined problem. In parallel, Action Research (AR) generates knowledge through collaborative, iterative cycles of planned action and critical reflection conducted with stakeholders in real settings. Bringing both traditions together, Action Design Research (ADR) integrates DSR and AR by developing and evaluating artifacts through participatory cycles focused on intervention while maintaining explicit expectations of rigor and contribution. These approaches are often used in international or study abroad research contexts, which are research environments spanning national, cultural, institutional, or governance boundaries and therefore require adaptive methods, careful ethical attention, and sustained stakeholder engagement. This synthesis results in an integrated methodological framework that positions Action Design Research as a supervisory scaffold for doctoral training in applied computing and information systems. The framework integrates Design Science Research and Action Research within an iterative cycle embedded in dialogical supervision and ethical reflexivity. It contributes a structured model for aligning methodological rigor, doctoral learning, and practical impact in complex and international research environments. Full article

Crop residue open burning is a major source of atmospheric pollutants that degrade regional air quality, enhance climate forcing, and threaten public health through emissions of particulate matter, greenhouse gases, and toxic species. In southern China, satellite-based emission estimates are often underestimated because frequent cloud cover and limited spatiotemporal resolution hinder the detection of agricultural fires. In this study, crop residue open burning emissions in Guangxi province from 2017 to 2023 were quantified using a statistical approach. The open burning proportion (OBP) was updated on an annual basis using the Visible Infrared Imaging Radiometer Suite (VIIRS) 375 m active fire product (VNP14IMG), and recently reported emission factors (EF_S) were adopted to enhance estimation accuracy. Annual emissions of pollutants were then spatially distributed to 0.05° × 0.05° grid cells based on satellite-detected fire counts and land cover information. The results indicated the total emissions of black carbon (BC), organic carbon (OC), sulfur dioxide (SO₂), nitric oxide (NO_X), carbon monoxide (CO), carbon dioxide (CO₂), fine particles (PM_2.5), coarse particles (PM₁₀), ammonia (NH₃), methane (CH₄) and non-methane volatile organic compound (NMVOC) in Guangxi province during 2017–2023 were 58.90, 230.48, 37.90, 213.95, 4234.41, 108,775.48, 583.09, 667.70, 46.36, 322.74 and 710.20 Gg, respectively. Sugarcane residue burning was identified as the dominant contributor, accounting for 41.26–64.38% of total emissions, followed by rice (20.66–43.06%), corn (5.11–17.25%), and cassava (4.33–6.45%). Emissions exhibited clear interannual variability, declining from 2017 to 2020 under strict control measures and increasing again from 2021 to 2023 as enforcement weakened. Incorporating annually updated VIIRS-derived OBP_S into the statistical inventory improves the temporal representation and reliability of multi-year emission estimates for agricultural burning. Full article

Deep neural networks suffer performance drops when source and target domains differ in distribution, motivating research into domain adaptation (DA). Traditional DA approaches presume source samples come from a single domain and can be available during adaptation. Nevertheless, in real-world applications, multiple source domains often exist, and source samples may be inaccessible owing to privacy and storage limitations. In response to the challenges of multi-source and source-free, multi-source-free domain adaptation (MSFDA) is proposed, which captures transferable information from a set of pre-trained source models to boost performance of the model on target domain. Most MSFDA methods meet these challenges by utilizing pseudo-labeling. However, pseudo-labels generated by distinct source models may contain noise and even be contradictory, which weakens their efficacy in facilitating source models adapting to the target domain. Moreover, these methods do not consider class imbalance, which would lead to biased predictions for minority classes, and undermine adaptation. Therefore, we propose a novel MSFDA method which extends adversarial learning to a multi-source-free setting. This method presents proxy multi-source domain adversarial learning, which aligns target features extracted by different source models in an adversarial manner, enhancing the capability of source models to extract domain-invariant features and potentially obtain high-quality pseudo-labels. Moreover, a nuclear-norm maximization regularization is employed to constrain prediction matrices, which can reduce the prediction uncertainty and enhance the discriminability of the model, while mitigating the prediction bias and promoting the prediction accuracy for minority classes. Finally, comprehensive evaluations on four benchmark datasets prove the validity of the proposed method. Full article