Network Data Research Articles

Understanding College Students' Help-Seeking Intentions for Alcohol Use: A Theoretical and Network-Based Approach.

This study investigated the association between college students' social network characteristics, the RAA, and intentions to seek help for alcohol use if they thought or knew they had a problem. Cross-sectional survey data from 1,447 college students were extracted from a larger, nationally representative college student quota sample. Inclusion criteria were between 18 to 24 years of age, enrolled full time in a college or university, reported past-year alcohol consumption, and passed data quality checks. Participants reported demographics, RAA constructs, and egocentric network data. Network composition variables were calculated via E-Net. Descriptive statistics and hierarchical linear regression analyses were conducted to examine intentions to seek help. A regression model consisting of RAA constructs was statistically significant at the p < .001 level, accounting for nearly 40% of the variance in help-seeking intentions. The addition of network composition variables increased the explained variance by an additional 2.8%, which was statistically significant at the p < .001 level. Integrating RAA constructs with social network characteristics can help identify specific individual beliefs and social influences that can be targeted to strengthen help-seeking intentions and behaviors among college students who engage in high-risk drinking.

Drainage Network Generation for Urban Pluvial Flooding (UPF) Using Generative Adversarial Networks (GANs) and GIS Data

Mapping urban pluvial flooding (UPF) in data-scarce regions poses significant challenges, particularly when drainage systems are inadequate or outdated. These limitations hinder effective flood mitigation and risk assessment. This study proposes an innovative approach to address these challenges by integrating deep learning (DL) models with traditional methods. First, deep convolutional generative adversarial networks (DCGANs) were employed to enhance drainage network data generation. Second, deep recurrent neural networks (DRNNs) and multi-criteria decision analysis (MCDA) methods were implemented to assess UPF. The study compared the performance of these approaches, highlighting the potential of DL models in providing more accurate and robust flood mapping outcomes. The methodology was applied to Lahore, Pakistan—a rapidly urbanizing and data-scarce region frequently impacted by UPF during monsoons. High-resolution ALOS PALSAR DEM data were utilized to extract natural drainage networks, while synthetic datasets generated by GANs addressed the lack of historical flood data. Results demonstrated the superiority of DL-based approaches over traditional MCDA methods, showcasing their potential for broader applicability in similar regions worldwide. This research emphasizes the role of DL models in advancing urban flood mapping, providing valuable insights for urban planners and policymakers to mitigate flooding risks and improve resilience in vulnerable regions.

Advances to IoT security using a GRU-CNN deep learning model trained on SUCMO algorithm

The rapid expansion of the Internet of Things (IoT) has significantly improved the various aspects of our daily life. However, along with its benefits, new security threats such as Denial of Service (DoS) attacks and Botnets have emerged. To adopt this technology and integrity of IoT environment, detection of such attacks become crucial. This paper proposes a hybrid deep learning model that combines Convolutional Neural Network (CNN) and Gated Recurrent Units (GRUs) to classify the IoT security threats. The CNN is used to extract the spatial features from the network data, where on the other hand GRUs used for capturing the temporal dependencies. This combination makes the model effective at analysing both static and dynamic aspects of network data. Further, to optimize the performance of the proposed hybrid model, self-upgraded Cat and Mouse Optimization (SUCMO) algorithm is employed, a state of art optimization technique. The SUCMO algorithm fine-tunes the deep learning model’s hyperparameters to improve classification accuracy. The proposed model is evaluated through experiments on two different datasets i.e., UNSW-NB15 and BoT-IoT, and results demonstrates that proposed work outperforms the traditional work as well as state of the art works.

A quality enhancement method for vehicle trajectory data using onboard images

ABSTRACT Crowdsourcing is essential for real-time map updates. However, low-cost GPS devices often cause issues such as trajectory point deviation and key attribute missing, which reduces the potential application of trajectory data and subsequently impacts the accuracy of map updates and the quality of transportation services. Map-matching algorithms (MMs) are commonly employed to vertically match trajectory points with the road on which the vehicle is traveling, based on characteristics between the trajectory points and urban road network data. Nonetheless, challenges such as relative deviation between trajectory points, overall deviation parallel to the road direction, and abnormal vehicle heading angles still persist, increasing the difficulty of trajectory position correction. Considering the potential of onboard videos in crowdsourced data to reflect the driving environment and attitude changes of vehicles, this study proposes a trajectory quality enhancement method using onboard videos and computer vision technology to address the aforementioned issues. The experimental results indicate that the maximum difference between the estimated angle and the true angle does not exceed 35°. The estimated position can correct over 95% of the local offsets in trajectory points, and the accuracy of trajectory overall positional deviation correction is within 20 m. These findings demonstrate that the proposed method effectively addresses the impacts of angle abnormalities, trajectory point local deviations, and trajectory overall positional deviations on MMs. Furthermore, integrating the proposed method with MMs can significantly enhance trajectory position correction accuracy, which provides valuable technical support and method references for promoting the practical application of map crowdsourcing updates.

Assessing equity of care across metastatic breast cancer treatment junctures: a multi-site retrospective cohort study

BackgroundAlthough inequality in the prevalence and mortality of metastatic breast cancer (mBC) have been documented, the contribution of non-clinical predictors along the care continuum remains unclear. This study aimed to identify determinants of disparities along the continuum of care among patients with mBC.MethodsA retrospective cohort study of patients with mBC under the Medicare Oncology Care Model program was conducted with Texas Oncology Network data from July 2016 to June 2021 as well as the County Health Rankings and Roadmaps Data. Study outcomes were rates of completion and time to receipt of services across four key care junctures: (1) Diagnosis and Evaluation, (2) Treatment Plan Design, (3) Treatment Implementation, and (4) End-of-life Care. Race/ethnicity was the primary predictor variable and secondary predictors included sociodemographic and clinical characteristics (referred to as the Vulnerability Cluster). Bivariate analysis, multivariable logistic regression, and generalized linear models were employed.ResultsAmong 460 patients included, mean age was 72.7 years and 98.7% were female; 73.7% were white, followed by 10.7% Hispanic, and 7.6% Black. Over 90% received services across the first three junctures. At End-of-life Care, only 14 (3.04%) received palliative care referral and 34.1% were enrolled in hospice, with an average length of stay (LOS) of 32.0 days. Adjusted analyses revealed significant racial/ethnic disparities in the End-of-life Care juncture, especially in hospice utilization. Compared to whites, blacks and Hispanics had shorter days to hospice enrollment, by 13.2% (Incidence Rate Ratio [IRR] = 0.868, 95% CI = 0.845–0.891, p < 0.001) and 34.8% (IRR = 0.652, 95% CI = 0.635–0.671, p < 0.001), respectively, and shorter LOS by 24.6 (IRR = 0.754, 95% CI = 0.645–0.882, p = 0.0004) and 25.3% (IRR = 0.747, 95% CI = 0.632–0.882, p = 0.0006), respectively. Other vulnerability cluster variables were predictive across care junctures.ConclusionsThe vulnerability cluster of variables, even in a Medicare population, were associated with the receipt and timeliness of care, especially at end-of-life among patients with mBC. More proactive and equitable care in mBC toward the end of their care journey is warranted.

InDepth: A Distributed Data Collection System for Modern Computer Networks

Cybersecurity researchers and security analysts rely heavily on data to train and test network threat detection models, and to conduct post-breach forensic analyses. Comprehensive data-including network traces, host telemetry, and contextual information-are crucial for these tasks. However, widely used public datasets often suffer from outdated network traffic and features, statistical anomalies, and simulation artifacts. Furthermore, existing data collection systems frequently face architectural and computational limitations, necessitating workarounds that result in incomplete or disconnected data. Currently, no framework provides comprehensive data collection from all network segments without requiring specialized or proprietary hardware or software agents. This paper introduces InDepth, a scalable system employing a distributed, data-link layer architecture that enables comprehensive data acquisition across entire networks. We also present a model cyber range capable of dynamically generating datasets for evaluation. We demonstrate the effectiveness of InDepth using real-world network data.

Research on multimodal English teaching model based on genetic algorithm

In recent years, English language teaching has undergone significant transformation due to the integration of technology and innovative teaching methodologies. Among the most promising advancements is the formation of multimodal teaching, which incorporates various forms of media. Despite the potential benefits associated with multimodal teaching, challenges remain in tailoring content to individual learning needs and optimizing the learning process for maximum effectiveness. To address these challenges, this research proposes a Multimodal English Teaching Framework (METF) to enhance the effectiveness and personalization of language learning. The framework integrates multiple teaching modalities, including text, audio, video, and interactive elements, to accommodate diverse learning preferences. It incorporates student interaction data (responses, time spent on tasks, and engagement levels), learner demographics, learning styles, and feedback records, thereby capturing the effectiveness of personalized content delivery. Pre-processing steps, including tokenization for text data, noise reduction for audio data, and normalization for resolution and interaction data, ensure data quality and consistency. Feature extraction techniques such as Convolutional Neural Networks (CNNs) for audio and video data and Term Frequency-Inverse Document Frequency (TF-IDF) for text data are applied to capture meaningful patterns. To enhance the fusion of multimodal data, feature-level fusion is implemented, integrating recovered characteristics from multiple modalities into a unified representation that supports more accurate decision-making. This research establishes the Resilient Genetic Algorithm (RGA) to enhance the adaptive learning process by ensuring robustness and avoiding premature convergence. Experimental results demonstrate the potential of this multimodal approach, achieving 94.2% accuracy, a convergence time of 45%, task engagement of 89.4%, and personalization effectiveness of 91.7%. These results indicate a high value of accuracy in multimodal English teaching. This research highlights the applicability of algorithms in optimizing complex, multimodal educational environments.

Finding Aids and Networked Biography: Stuart Hall and the Centre for Contemporary Cultural Studies

Finding aids – tools to help researchers locate and understand archival materials – are rich with biographical information: in archival descriptions, we find personal milestones stated alongside traceable genealogies of ideas and associations between people and institutions. A pilot project at the University of Birmingham tested digital methods for distilling this information as a means to life writing; this paper reports on its findings. The pilot project involved extracting networked data from finding aids associated with Stuart Hall (1932-2014) – a founding figure of cultural studies – and the Centre for Contemporary Cultural Studies (CCCS; 1964-2002). A tool was developed to capture data around the movement, meetings, connections, and output of individuals and groups. The project aimed to visualise Hall’s life (people and entities he worked with, events he attended, etcetera) and the institutional ‘life’ of the CCCS (flow of finance, members, and influences). This paper outlines challenges faced by the project team and steps taken to circumvent them, and considers matching the form of digital representation to the distinctive pedagogical approach advocated by Hall and the CCCS. With the particularity of this dataset in mind, this paper discusses how digital methods can be applied to extract and organise finding aid data as a means of framing Hall’s biography.

Infrared small target detection based on multi-scale dense network and deep data fusion paradigm

Infrared small target detection based on multi-scale dense network and deep data fusion paradigm

Integrating Human Mobility Models with Epidemic Modeling: A Framework for Generating Synthetic Temporal Contact Networks

High-resolution temporal contact networks are useful ingredients for realistic epidemic simulations. Existing solutions typically rely either on empirical studies that capture fine-grained interactions via Bluetooth or wearable sensors in confined settings or on large-scale simulation frameworks that model entire populations using generalized assumptions. However, for most realistic modeling of epidemic spread and the evaluation of countermeasures, there is a critical need for highly resolved, temporal contact networks that encompass multiple venues without sacrificing the intricate dynamics of real-world contacts. This paper presents an integrated approach for generating such networks by coupling Bayesian-optimized human mobility models (HuMMs) with a state-of-the-art epidemic simulation framework. Our primary contributions are twofold: First, we embed empirically calibrated HuMMs into an epidemic simulation environment to create a parameterizable, adaptive engine for producing synthetic, high-resolution, population-wide temporal contact network data. Second, we demonstrate through empirical evaluations that our generated networks exhibit realistic interaction structures and infection dynamics. In particular, our experiments reveal that while variations in population size do not affect the underlying network properties—a crucial feature for scalability—altering location capacities naturally influences local connectivity and epidemic outcomes. Additionally, sub-graph analyses confirm that different venue types display distinct network characteristics consistent with their real-world contact patterns. Overall, this integrated framework provides a scalable and empirically grounded method for epidemic simulation, offering a powerful tool for generating and simulating contact networks.

Water usage efficiency and productivity change: a non-convex metafrontier Malmquist index

This paper uses a non-convex metafrontier approach to study the temporal evolution and the productivity change in the water usage efficiency of 91 countries during the period 2015–2020. A circular three-stage Network Data Envelopment Analysis (NDEA) production process is considered. The non-radial Directional Distance Function (DDF) is used to compute the system efficiency as well as that of the three stages. Three sets of NDEA models are solved: contemporaneous group, intertemporal group and intertemporal metafrontier. The corresponding Malmquist index and its decomposition into efficiency change, best practice gap change and technology gap change is computed for each country and averaged for each world region. The results found indicate that water usage inefficiency seems to be limited to a relatively small number of countries and is due to shortfalls in the gross value added and treated municipal wastewater dimensions. The inefficiency in gross value added actually shows an upward trend. The inefficiency in the water withdrawal and treated municipal wastewater dimensions have remained approximately constant during this period while the inefficiency in produced municipal wastewater have decreased slightly. As regards productivity change, most regions had a negative evolution during this period. Central Asia and Middle East-Western Asia are the exception to this trend and their productivity improvement is mostly due to an improvement in their best practice gap and, to a lesser extent, in their technology gap ratios.

Mobile Network Data Security using Deep Learning

With the rapid expansion of mobile networks and the exponential growth of data transmission, ensuring robust security against cyber threats has become a critical challenge. Traditional security mechanisms struggle to keep pace with evolving attack strategies, necessitating intelligent and adaptive solutions. This research explores the application of deep learning techniques to enhance mobile network data security by detecting anomalies, preventing intrusions, and securing communication channels. The study leverages advanced neural network architectures, such as Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), and Transformer models, to analyze traffic patterns and identify potential threats in real time. A hybrid deep learning framework is proposed to optimize threat detection efficiency while minimizing false positives. Experimental evaluations on real-world datasets demonstrate significant improvements in accuracy, speed, and resilience against sophisticated cyberattacks compared to conventional security methods. The findings highlight the potential of deep learning-based approaches in strengthening mobile network security, ensuring data integrity, and safeguarding user privacy in an increasingly interconnected digital landscape.

Edge vs. Cloud: Empirical Insights into Data-Driven Condition Monitoring

This study evaluates edge and cloud computing paradigms in the context of data-driven condition monitoring of rotating electrical machines. Two well-known platforms, the Raspberry Pi and Amazon Web Services Elastic Compute Cloud, are used to compare and contrast these two computing paradigms in terms of different metrics associated with their application suitability. The tested induction machine fault diagnosis models are developed using popular algorithms, namely support vector machines, k-nearest neighbours, and decision trees. The findings reveal that while the cloud platform offers superior computational and memory resources, making it more suitable for complex machine learning tasks, it also incurs higher costs and latency. On the other hand, the edge platform excels in real-time processing and reduces network data burden, but its computational and memory resources are found to be a limitation with certain tasks. The study provides both quantitative and qualitative insights into the trade-offs involved in selecting the most suitable computing approach for condition monitoring applications. Although the scope of the empirical study is primarily limited to factors such as computational efficiency, scalability, and resource utilisation, particularly in the context of specific machine learning models, this paper offers broader discussion and future research directions of other key issues, including latency, network variability, and energy consumption.

Fault and Performance Management in SD-WAN: Overcoming Key Challenges

Software-Defined Wide Area Networks (SD-WAN) have transformed enterprise networking by separating network control from hardware infrastructure and implementing centralized management through software controllers. This transformative technology offers organizations unprecedented flexibility, cost efficiency, and deployment simplicity compared to legacy WAN architectures. However, new fault and performance management challenges have emerged as SD-WAN adoption accelerates alongside multi-cloud strategies and hybrid work models. The dynamic nature of SD-WAN environments featuring multiple transport options and intelligent traffic steering creates complexity that traditional monitoring approaches cannot adequately address. Organizations face difficulties with link congestion, suboptimal path selection, real-time latency fluctuations, and application identification, which can significantly impact user experience. Promising solutions leveraging artificial intelligence and machine learning technologies are emerging to address these challenges, providing capabilities for anomaly detection, predictive maintenance, event correlation, and adaptive monitoring. Optimization strategies, including intelligent traffic routing, dynamic bandwidth allocation, application-aware policies, and error correction techniques, enhance SD-WAN performance. As organizations increasingly adopt multi-cloud architectures, achieving end-to-end visibility becomes essential yet challenging, requiring integrated monitoring approaches that bridge traditional silos between network, application, and cloud performance data. By implementing comprehensive performance management strategies and leveraging advanced technologies, organizations can maximize the benefits of their SD-WAN deployments while ensuring reliable connectivity and optimal application performance across distributed environments.

A Real Network Environment Dataset for Traffic Analysis

The objective of internet traffic analysis is to identify latent patterns and ascertain the true state of internet operations by examining traffic data. This approach is considered an effective and valuable means to achieve accurate network management. Whilst the extant network traffic datasets are predominantly collated within a laboratory environment, exhibiting deficiencies with regard to authenticity in terms of network scales, users, behaviours, and temporal and spatial characteristics, this paper proposes an in-situ network deployment and data collection scheme involving a large number of devices and users. The scheme involves the collection of a large real Internet traffic dataset including encrypted and non-encrypted traffic through sensors deployed on real-world network access equipment. Through desensitization, cleaning, feature engineering and labelling, an open database is created for researchers in the field of traffic analysis to use in academic and engineering.

Spatial Mismatch Between Transportation Development and Tourism Spatial Vitality in Yunnan Province in the Context of Urban–Rural Integration

As China’s urban–rural integration progresses, the connections between urban and rural areas continue to strengthen, making the spatial matching between transportation infrastructure and tourism resources increasingly crucial for coordinated regional development. This study investigates the spatial–temporal mismatch between transportation development and tourism spatial vitality in Yunnan Province, proposing optimization strategies to improve their coordination. Using Weibo check-in big data and OpenStreetMap transportation network data, we apply Convolutional Long Short-Term Memory (ConvLSTM) networks and bivariate spatial autocorrelation analysis to examine this relationship. The results show strong transportation–tourism matching in Kunming and surrounding areas. However, northwest and southern Yunnan exhibit significant mismatches—despite transportation improvements, underdeveloped tourism resources constrain vitality growth. Particularly in some remote regions, well-developed transportation infrastructure coexists with low tourism vitality, revealing persistent spatial mismatches between transport facilities and tourism resources. In general, transportation infrastructure development generally enhances tourism spatial vitality, but requires coordinated tourism resource development and market demand alignment. The study results provide a basis for improving the coordinated development of transportation and tourism, offering practical guidance for policymakers to promote balanced regional development and urban–rural integration.

A Review of Advances in Probable Maximum Precipitation and Flood Estimation in the Kaligandaki Basin, Nepal

Abstract The Kaligandaki River Basin in Nepal, a region characterized by extreme precipitation variability and frequent flooding, has been the focus of recent hydrological studies to estimate Probable Maximum Precipitation (PMP) and Probable Maximum Flood (PMF). This review synthesizes the methodologies, findings, and challenges of a pivotal study that employed the HEC-HMS model to estimate PMF from PMP in the Kaligandaki Basin. The study’s approach, which integrates hydro-meteorological data, statistical methods, and hydrological modeling, is evaluated in the context of global PMP/PMF estimation practices. Key strengths include robust data utilization and sub-basin-specific analysis, while limitations, such as sparse data networks and the omission of climate change impacts, are highlighted. Recommendations for advancing PMP/PMF research in data-scarce, topographically complex regions, such as Nepal, are provided, emphasizing the need for enhanced data collection, multi-model approaches, and climate-resilient methodologies.The Kaligandaki River Basin study represents a significant advancement in hydrological research for Nepal's complex terrain. By employing the HEC-HMS model to estimate Probable Maximum Flood (PMF) from Probable Maximum Precipitation (PMP), the research provides crucial insights into extreme flood scenarios in a region prone to hydrological extremes. The methodology's strength lies in its comprehensive approach, integrating diverse hydro-meteorological data and statistical methods to produce sub-basin-specific analyses. This granular approach allows for a more nuanced understanding of flood risks across the basin's varied topography. However, the study faces notable challenges inherent to its geographical context. The sparse data networks in Nepal limit the accuracy and spatial resolution of the analysis, potentially affecting the reliability of PMP and PMF estimates. Additionally, the omission of climate change impacts in the modeling process represents a significant gap, given the increasing influence of global warming on precipitation patterns and flood frequencies. To address these limitations and enhance future research, there is a pressing need for improved data collection infrastructure, the adoption of multi-model approaches to reduce uncertainty, and the integration of climate change projections into PMP/PMF estimations. These advancements would not only improve the accuracy of flood risk assessments but also contribute to more effective flood management and climate adaptation strategies in the Kaligandaki Basin and similar data-scarce, topographically complex regions worldwide. Keywords: Kaligandaki River Basin, Probable Maximum Precipitation (PMP), Probable Maximum Flood (PMF), HEC-HMS model, Hydrological modeling, Flood risk assessment, Nepal hydrology, Extreme precipitation, Data-scarce regions, Climate change impacts, Topographically complex terrain

Gender-based differences in eco-efficient farming

This study investigates gender-based differences in eco-efficiency among Hungarian field crop farms, using data from the Farm Accountancy Data Network between 2015 and 2020. Applying Data Envelopment Analysis and Blinder-Oaxaca decompositions, we reveal a consistent eco-efficiency advantage for women-headed farms, particularly at mid-quantiles of eco-efficiency distribution. This advantage is largely attributed to women’s ability to optimize resource use effectively, though unexplained factors also contribute, suggesting potential differences in management practices. Results highlight that women-led farms often adopt eco-efficient practices that may contribute to sustainability goals. These findings highlight the need for policies that support women farmers’ access to resources, knowledge, and innovation in eco-friendly farming practices, helping to enhance sustainability in agricultural production. Encouraging women participation in sustainable agricultural entrepreneurship could play a critical role in driving eco-efficiency across the crop sector, ultimately contributing to environmental resilience and rural socio-economic development.

Multiscale Two-Stream Fusion Network for Benggang Classification in Multi-Source Images.

Benggangs, a type of soil erosion widely distributed in the hilly and mountainous regions of South China, pose significant challenges to land management and ecological conservation. Accurate identification and assessment of their location and scale are essential for effective Benggang control. With advancements in technology, deep learning has emerged as a critical tool for Benggang classification. However, selecting suitable feature extraction and fusion methods for multi-source image data remains a significant challenge. This study proposes a Benggang classification method based on multiscale features and a two-stream fusion network (MS-TSFN). Key features of targeted Benggang areas, such as slope, aspect, curvature, hill shade, and edge, were extracted from Digital Orthophotography Map (DOM) and Digital Surface Model (DSM) data collected by drones. The two-stream fusion network, with ResNeSt as the backbone, extracted multiscale features from multi-source images and an attention-based feature fusion block was developed to explore complementary associations among features and achieve deep fusion of information across data types. A decision fusion block was employed for global prediction to classify areas as Benggang or non-Benggang. Experimental comparisons of different data inputs and network models revealed that the proposed method outperformed current state-of-the-art approaches in extracting spatial features and textures of Benggangs. The best results were obtained using a combination of DOM data, Canny edge detection, and DSM features in multi-source images. Specifically, the proposed model achieved an accuracy of 92.76%, a precision of 85.00%, a recall of 77.27%, and an F1-score of 0.8059, demonstrating its adaptability and high identification accuracy under complex terrain conditions.

Sensor network- and frequency-based damage identification approach in multi-damage structures

This paper presents a novel Sensor Network and Frequency-Based (SNFB) Structural Health Monitoring (SHM) approach, designed to identify damage locations in multi-damaged structures. The SNFB method, a model-based SHM technique, utilizes the Frequency Response Function (FRF). Unlike existing FRF-based methods, which rely on factors such as sensor number and placement, damage number, intensity, location, and frequency range, the proposed SNFB method overcomes these limitations. It integrates hypothesis testing theory and sensor network data fusion to establish an accurate threshold for detecting multiple damages. Experimental tests and numerical models were adopted on beams equipped with sensor networks, considering various damage scenarios. By comparing SNFB performance to conventional FRF-based methods, the study highlights its benefits. Results demonstrate that the SNFB approach accurately detects damage locations in multi-damaged beams. Moreover, it was shown that SNFB method identifies the damages that result in a reduction of at least 5% in the stiffness of the structure. This advancement in FRF-based technique paves the way to make it applicable in real-world long structures, which have several damages with different intensities.