Search Results (1,770)

Archaeological research fundamentally relies on detecting features to uncover hidden historical information. Airborne (aerial) LiDAR technology has significantly advanced this field by providing high-resolution 3D terrain maps that enable the identification of ancient structures and landscapes with improved accuracy and efficiency. This technical note comprehensively reviews 45 recent studies to critically examine the integration of Machine Learning (ML) and Deep Learning (DL) techniques, particularly Convolutional Neural Networks (CNNs), with airborne LiDAR derivatives for automated archaeological feature detection. The review highlights the transformative potential of these approaches, revealing their capability to automate feature detection and classification, thus enhancing efficiency and accuracy in archaeological research. CNN-based methods, employed in 32 of the reviewed studies, consistently demonstrate high accuracy across diverse archaeological features. For example, ancient city walls were delineated with 94.12% precision using U-Net, Maya settlements with 95% accuracy using VGG-19, and with an IoU of around 80% using YOLOv8, and shipwrecks with a 92% F1-score using YOLOv3 aided by transfer learning. Furthermore, traditional ML techniques like random forest proved effective in tasks such as identifying burial mounds with 96% accuracy and ancient canals. Despite these significant advancements, the application of ML/DL in archaeology faces critical challenges, including the scarcity of large, labeled archaeological datasets, the prevalence of false positives due to morphological similarities with natural or modern features, and the lack of standardized evaluation metrics across studies. This note underscores the transformative potential of LiDAR and ML/DL integration and emphasizes the crucial need for continued interdisciplinary collaboration to address these limitations and advance the preservation of cultural heritage. Full article

Existing approaches to social media sentiment analysis typically focus on static classification, offering limited foresight into how public opinion evolves. This study addresses that gap by introducing the Multi-Feature Sentiment-Driven Forecasting (MFSF) framework, a novel pipeline that enhances sentiment trend prediction by integrating rich contextual information from text. Using state-of-the-art transformer models on the Sentiment140 dataset, our framework extracts three concurrent signals from each tweet: sentiment polarity, aspect-based scores (e.g., ‘price’ and ‘service’), and topic embeddings. These features are aggregated into a daily multivariate time series. We then employ a SARIMAX model to forecast future sentiment, using the extracted aspect and topic data as predictive exogenous variables. Our results, validated on the historical Sentiment140 Twitter dataset, demonstrate the framework’s superior performance. The proposed multivariate model achieved a 26.6% improvement in forecasting accuracy (RMSE) over a traditional univariate ARIMA baseline. The analysis confirmed that conversational aspects like ‘service’ and ‘quality’ are statistically significant predictors of future sentiment. By leveraging the contextual drivers of conversation, the MFSF framework provides a more accurate and interpretable tool for businesses and policymakers to proactively monitor and anticipate shifts in public opinion. Full article

Governments worldwide have set ambitious targets for decarbonising energy grids, driving the need for increased renewable energy generation and improved energy efficiency. One key strategy for achieving this involves enhanced energy management in buildings, often using machine learning-based forecasting methods. However, such methods typically rely on extensive historical data collected via costly sensor installations—resources that many buildings lack. This study introduces a novel forecasting approach that eliminates the need for large-scale historical datasets or expensive sensors. By integrating custom-built models with existing energy data, the method applies calculated weighting through a distance matrix and accuracy coefficients to generate reliable forecasts. It uses readily available building attributes—such as floor area and functional type to position a new building within the matrix of existing data. A Euclidian distance matrix, akin to a K-nearest neighbour algorithm, determines the appropriate neural network(s) to utilise. These findings are benchmarked against a consolidated, more sophisticated neural network and a long short-term memory neural network. The dataset has hourly granularity over a 24 h horizon. The model consists of five bespoke neural networks, demonstrating the superiority of other models with a 610 s training duration, uses 500 kB of storage, achieves an R² of 0.9, and attains an average forecasting accuracy of 85.12% in predicting the energy consumption of the five buildings studied. This approach not only contributes to the specific goal of a fully decarbonized energy grid by 2050 but also establishes a robust and efficient methodology for maintaining standards with existing benchmarks while providing more control over the method. Full article

Multi-behavior sequential recommendation (MBSRec) is a form of sequential recommendation. It leverages users’ historical interaction behavior types to better predict their next actions. This approach fits real-world scenarios better than traditional models do. With the rise of the transformer model, attention mechanisms are widely used in recommendation algorithms. However, they suffer from low-pass filtering, and the simple learnable positional encodings in existing models offer limited performance gains. To address these problems, we introduce the context-aware multi-behavior sequential recommendation model (CAMBSRec). It separately encodes items and behavior types, replaces traditional positional encoding with context-similarity positional encoding, and applies the discrete Fourier transform to separate the high and low frequency components and enhance the high frequency components, countering the low-pass filtering effect. Experiments on three public datasets show that CAMBSRec performs better than five baseline models, demonstrating its advantages in terms of recommendation performance. Full article

Concrete mix design plays a pivotal role in ensuring the mechanical performance, durability, and sustainability of construction projects. However, the nonlinear interactions among the mix components challenge traditional approaches in predicting compressive strength and optimizing proportions. This study presents a two-stage hybrid framework that integrates deep learning with reinforcement learning to overcome these limitations. First, a Convolutional Neural Network–Long Short-Term Memory (CNN–LSTM) model was developed to capture spatial–temporal patterns from a dataset of 1030 historical concrete samples. The extracted features were enhanced using an eXtreme Gradient Boosting (XGBoost) meta-model to improve generalizability and noise resistance. Then, a Dueling Double Deep Q-Network (Dueling DDQN) agent was used to iteratively identify optimal mix ratios that maximize the predicted compressive strength. The proposed framework outperformed ten benchmark models, achieving an MAE of 2.97, RMSE of 4.08, and R² of 0.94. Feature attribution methods—including SHapley Additive exPlanations (SHAP), Elasticity-Based Feature Importance (EFI), and Permutation Feature Importance (PFI)—highlighted the dominant influence of cement content and curing age, as well as revealing non-intuitive effects such as the compensatory role of superplasticizers in low-water mixtures. These findings demonstrate the potential of the proposed approach to support intelligent concrete mix design and real-time optimization in smart construction environments. Full article

Next Point-of-Interest (POI) recommendation is aimed at predicting users’ future visits based on their current status and historical check-in records, providing convenience to users and potential profits to businesses. The Graph Neural Network (GNN) has become a common approach for this task due to the capabilities of modeling relations between nodes in a global perspective. However, most existing studies overlook the more prevalent heterogeneous relations in real-world scenarios, and manually constructed graphs may suffer from inaccuracies. To address these limitations, we propose a model called Heterogeneous Graph Structure Learning for Next POI Recommendation (HGSL-POI), which integrates three key components: heterogeneous graph contrastive learning, graph structure learning, and sequence modeling. The model first employs meta-path-based subgraphs and the user–POI interaction graph to obtain initial representations of users and POIs. Based on these representations, it reconstructs the subgraphs through graph structure learning. Finally, based on the embeddings from the reconstructed graphs, sequence modeling incorporating graph neural networks captures users’ sequential preferences to make recommendations. Experimental results on real-world datasets demonstrate the effectiveness of the proposed model. Additional studies confirm its robustness and superior performance across diverse recommendation tasks. Full article

This study presents a cross-cultural comparison of two fortified residences—Palazzo Farnese in Italy and Dong’s Fortified Compound in China—through a triadic analytical framework encompassing architectural form, symbolic ornamentation, and public perception. By combining field observation, iconographic interpretation, and digital ethnography, the research investigates how heritage meaning is constructed, encoded, and reinterpreted across distinct sociocultural contexts. Empirical materials include architectural documentation, decorative analysis, and a curated dataset of 4947 user-generated images and 1467 textual comments collected from Chinese and international platforms between 2020 and 2024. Methods such as CLIP-based visual clustering and BERTopic-enabled sentiment modelling were applied to extract patterns of perception and symbolic emphasis. The findings reveal contrasting representational logics: Palazzo Farnese encodes dynastic authority and Renaissance cosmology through geometric order and immersive frescoes, while Dong’s Compound conveys Confucian ethics and frontier identity via nested courtyards and traditional ornamentation. Digital responses diverge accordingly: international users highlight formal aesthetics and photogenic elements; Chinese users engage with symbolic motifs, family memory, and ritual significance. This study illustrates how historically fortified residences are reinterpreted through culturally specific digital practices, offering an interdisciplinary approach that bridges architectural history, symbolic analysis, and digital heritage studies. Full article

This study presents the first systematic empirical comparison of neural architectures for Universal Dependencies (UD) parsing in Old English, thus addressing central questions in computational historical linguistics and low-resource language processing. We evaluate three approaches—a baseline spaCy pipeline, a pipeline with a pretrained tok2vec component, and a MobileBERT transformer-based model—across datasets ranging from 1000 to 20,000 words. Our results demonstrate that the pretrained tok2vec model consistently outperforms alternatives, because it achieves 83.24% UAS and 74.23% LAS with the largest dataset, whereas the transformer-based approach substantially underperforms despite higher computational costs. Performance analysis reveals that basic tagging tasks reach 85–90% accuracy, while dependency parsing achieves approximately 75% accuracy. We identify critical scaling thresholds, with substantial improvements occurring between 1000 and 5000 words and diminishing returns beyond 10,000 words, which provides insights into scaling laws for historical languages. Technical analysis reveals that the poor performance of the transformer stems from parameter-to-data ratio mismatches (1250:1) and the unique orthographic and morphological characteristics of Old English. These findings defy assumptions about transformer superiority in low-resource scenarios and establish evidence-based guidelines for researchers working with historical languages. The broader significance of this study extends to enabling an automated analysis of three million words of extant Old English texts and providing a framework for optimal architecture selection in data-constrained environments. Our results suggest that medium-complexity architectures with monolingual pretraining offer superior cost–benefit trade-offs compared to complex transformer models for historical language processing. Full article

Structural cracks significantly threaten the safety and longevity of historical buildings, which are essential parts of cultural heritage. Conventional inspection techniques, which depend heavily on manual visual evaluations, tend to be inefficient and subjective. This research introduces an automated framework for crack and damage detection using advanced YOLO (You Only Look Once) models, aiming to improve both the accuracy and efficiency of monitoring heritage structures. A dataset comprising 2500 high-resolution images was gathered from historical buildings and categorized into four levels of damage: no damage, minor, moderate, and severe. Following preprocessing and data augmentation, a total of 5000 labeled images were utilized to train and evaluate four YOLO variants: YOLOv5, YOLOv8, YOLOv10, and YOLOv11. The models’ performances were measured using metrics such as precision, recall, mAP@50, mAP@50–95, as well as losses related to bounding box regression, classification, and distribution. Experimental findings reveal that YOLOv10 surpasses other models in multi-target detection and identifying minor damage, achieving higher localization accuracy and faster inference speeds. YOLOv8 and YOLOv11 demonstrate consistent performance and strong adaptability, whereas YOLOv5 converges rapidly but shows weaker validation results. Further testing confirms YOLOv10’s effectiveness across different structural components, including walls, beams, and ceilings. This study highlights the practicality of deep learning-based crack detection methods for preserving building heritage. Future advancements could include combining semantic segmentation networks (e.g., U-Net) with attention mechanisms to further refine detection accuracy in complex scenarios. Full article

The morphological diversity of Moroccan honey bees (Apis mellifera) was investigated using geometric morphometrics to assess wing venation patterns among three populations representing three climatic zones: desert, semiarid, and Mediterranean. A total of 193 honey bee samples were analyzed and compared to historical reference samples from the Morphometric Bee Data Bank in Oberursel, representing the three subspecies: A. m. intermissa, A. m. sahariensis, and A. m. major. Principal component analysis and linear discriminant analysis revealed significant, yet overlapping morphological differences among the climatic groups. Spatial modeling showed a significant southwest–northeast clinal gradient in wing morphology. Almost all samples were assigned to the African evolutionary lineage, except one, suggesting a dominant African genetic background. Interestingly, all three populations showed greater morphological affinity to A. m. intermissa than to A. m. sahariensis, which could indicate introgression or limitations in the current reference dataset. These discrepancies highlight the necessity of revising subspecies boundaries using updated morphometric and genomic approaches. These findings improve our understanding of honey bee biodiversity in Morocco and provide valuable information for conservation and breeding programs. Full article

Vessel trajectory prediction is fundamental to maritime navigation, safety, and operational efficiency, particularly as the industry increasingly relies on digital solutions and real-time data analytics. This study addresses the challenge of forecasting vessel movements using historical Automatic Identification System (AIS) data, with a focus on understanding the temporal behavior of deep learning models under different input and prediction horizons. To this end, a robust data pre-processing pipeline was developed to ensure temporal consistency, filter anomalous records, and segment continuous vessel trajectories. Using a curated dataset from the eastern Mediterranean, three deep recurrent neural network architectures, namely LSTM, Bi-LSTM, and Bi-GRU, were evaluated for short- and long-term trajectory prediction. Empirical results demonstrate that Bi-LSTM consistently achieves higher accuracy across both horizons, with performance gradually degrading under extended forecast windows. The analysis also reveals key insights into the trade-offs between model complexity, horizon-specific robustness, and predictive stability. This work contributes to maritime informatics by offering a comparative evaluation of recurrent architectures and providing a structured and empirical foundation for selecting and deploying trajectory forecasting models in operational contexts. Full article

The rapid electrification of transportation and the proliferation of rooftop solar photovoltaics (PVs) in urban environments are reshaping the operational dynamics of power distribution networks. However, the inherent uncertainty in electric vehicle (EV) behavior—including arrival times, charging preferences, and state-of-charge—as well as spatially and temporally variable solar generation, presents a profound challenge to existing scheduling frameworks. This paper proposes a novel data-driven distributionally robust optimization (DDRO) framework for solar-powered EV charging coordination under spatiotemporal uncertainty. Leveraging empirical datasets of EV usage and solar irradiance from a smart city deployment, the framework constructs Wasserstein ambiguity sets around historical distributions, enabling worst-case-aware decision-making without requiring the assumption of probability laws. The problem is formulated as a two-stage optimization model. The first stage determines day-ahead charging schedules, solar utilization levels, and grid allocations across an urban-scale distribution feeder. The second stage models real-time recourse actions—such as dynamic curtailment or demand reshaping—after uncertainties are realized. Physical grid constraints are modeled using convexified LinDistFlow equations, while EV behavior is segmented into user classes with individualized uncertainty structures. The model is evaluated on a modified IEEE 123-bus feeder with 52 EV-PV nodes, using 15 min resolution over a 24 h horizon and 12 months of real-world data. Comparative results demonstrate that the proposed DDRO method reduces total operational costs by up to 15%, eliminates voltage violations entirely, and improves EV service satisfaction by more than 30% relative to deterministic and stochastic baselines. This work makes three primary contributions: it introduces a robust, tractable optimization architecture that captures spatiotemporal uncertainty using empirical Wasserstein sets; it integrates behavioral and physical modeling within a unified dispatch framework for urban energy-mobility systems; and it demonstrates the value of robust coordination in simultaneously improving grid resilience, renewable utilization, and EV user satisfaction. The results offer practical insights for city-scale planners seeking to enable the reliable and efficient electrification of mobility infrastructure under uncertainty. Full article

Real-time quality monitoring using molten pool images is a critical focus in researching high-quality, intelligent automated welding. To address interference problems in molten pool images under complex welding scenarios (e.g., reflected laser spots from spatter misclassified as porosity defects) and the limited interpretability of deep learning models, this paper proposes a multi-granularity spatiotemporal representation learning algorithm based on the hybrid enhancement of handcrafted and deep learning features. A MobileNetV2 backbone network integrated with a Temporal Shift Module (TSM) is designed to progressively capture the short-term dynamic features of the molten pool and integrate temporal information across both low-level and high-level features. A multi-granularity attention-based feature aggregation module is developed to select key interference-free frames using cross-frame attention, generate multi-granularity features via grouped pooling, and apply the Convolutional Block Attention Module (CBAM) at each granularity level. Finally, these multi-granularity spatiotemporal features are adaptively fused. Meanwhile, an independent branch utilizes the Histogram of Oriented Gradient (HOG) and Scale-Invariant Feature Transform (SIFT) features to extract long-term spatial structural information from historical edge images, enhancing the model’s interpretability. The proposed method achieves an accuracy of 99.187% on a self-constructed dataset. Additionally, it attains a real-time inference speed of 20.983 ms per sample on a hardware platform equipped with an Intel i9-12900H CPU and an RTX 3060 GPU, thus effectively balancing accuracy, speed, and interpretability. Full article

Quantifying future changes in extreme events and associated flooding is challenging yet fundamental for stormwater managers. Along the U.S. Atlantic Coast, Eastern North Carolina (ENC) is frequently exposed to catastrophic floods from extreme rainfall that is typically associated with tropical cyclones. This study presents a novel approach that uses rainfall data from five dynamically and statistically downscaled (DD and SD) global climate models under two scenarios to visualize a potential future extent of flooding in ENC. Here, we use DD data (at 36-km grid spacing) to compute future changes in precipitation intensity–duration–frequency (PIDF) curves at the end of the 21st century. These PIDF curves are further applied to observed rainfall from Hurricane Matthew—a landfalling storm that created widespread flooding across ENC in 2016—to project versions of “Matthew 2100” that reflect changes in extreme precipitation under those scenarios. Each Matthew-2100 rainfall distribution was then used in hydrologic models (HEC-HMS and HEC-RAS) to simulate “2100” discharges and flooding extents in the Neuse River Basin (4686 km²) in ENC. The results show that DD datasets better represented historical changes in extreme rainfall than SD datasets. The projected changes in ENC rainfall (up to 112%) exceed values published for the U.S. but do not exceed historical values. The peak discharges for Matthew-2100 could increase by 23–69%, with 0.4–3 m increases in water surface elevation and 8–57% increases in flooded area. The projected increases in flooding would threaten people, ecosystems, agriculture, infrastructure, and the economy throughout ENC. Full article

The study relies on an extensive dataset (≈1.86 million news headlines) to investigate the heterogeneity and predictive power of explicit sentiment signals (TextBlob, VADER, and FinBERT) and implied sentiment (VIX) for stock market trends. We find that news content predominantly consists of objective or neutral information, with only a small portion carrying subjective or emotive weight. There is a structural market bias toward upswings (bullish market states). Market behavior appears anticipatory rather than reactive: forward-looking implied sentiment captures a substantial share (≈45–50%) of the variation in stock returns. By contrast, sentiment scores, even when disaggregated into firm- and non-firm-specific subscores, lack robust predictive power. However, weekend and holiday sentiment contains modest yet valuable market signals. Algorithm-wise, Gradient Boosting Machine (GBM) stands out in both classification (bullish vs. bearish) and regression tasks. Neither FinBERT news sentiment, historical returns, nor implied volatility offer a consistently exploitable edge over market efficiency. Thus, our findings lend empirical support to both the weak-form and semi-strong forms of the Efficient Market Hypothesis. In the realm of exploitable trading strategies, markets remain an enigma against systematic alpha. Full article