Utility Of Data Research Articles

DLPLSR: Dual Label Propagation-Driven Least Squares Regression with Feature Selection for Semi-Supervised Learning

In the real world, most data are unlabeled, which drives the development of semi-supervised learning (SSL). Among SSL methods, least squares regression (LSR) has attracted attention for its simplicity and efficiency. However, existing semi-supervised LSR approaches suffer from challenges such as the insufficient use of unlabeled data, low pseudo-label accuracy, and inefficient label propagation. To address these issues, this paper proposes dual label propagation-driven least squares regression with feature selection, named DLPLSR, which is a pseudo-label-free SSL framework. DLPLSR employs a fuzzy-graph-based clustering strategy to capture global relationships among all samples, and manifold regularization preserves local geometric consistency, so that it implements the dual label propagation mechanism for comprehensive utilization of unlabeled data. Meanwhile, a dual-feature selection mechanism is established by integrating orthogonal projection for maximizing feature information with an ℓ2,1-norm regularization for eliminating redundancy, thereby jointly enhancing the discriminative power. Benefiting from these two designs, DLPLSR boosts learning performance without pseudo-labeling. Finally, the objective function admits an efficient closed-form solution solvable via an alternating optimization strategy. Extensive experiments on multiple benchmark datasets show the superiority of DLPLSR compared to state-of-the-art LSR-based SSL methods.

Utilization of airborne geophysical data and remote sensing to identify radioactive and hydrothermal alteration zones in the East Qena area, Central Eastern Desert, Egypt

Abstract The central purpose of the present study is to pinpoint radioactive anomalies, unravel the pathways of uranium migration, map zones of hydrothermal alteration, and dissect the underlying geological structures within the East Qena area. This is achieved by applying airborne geophysical data and remote sensing techniques. The East Qena area was specifically chosen for its diverse rock units, encompassing basement rocks, especially granitic rocks, and crucial sedimentary formations like the phosphate and shale (Duwi, Dakhla, and Quseir formations). The airborne gamma-ray spectrometric data reveal radioactive zones with a relatively high content of potassium, uranium, and thorium. Ratio maps of eU/eTh and eU/K show significant uranium mineralization potential in the western part, particularly in the Esna, Dakhla, Duwi formations. It distinctly delineates the migrated uranium (U m) and F-parameter zones concentrated over the Duwi, Quseir, and Dakhla formations in the western part, as well as their presence over alkali feldspar granites in the eastern and northeastern parts of the area under study. Furthermore, remote sensing (RS) data have been instrumental in identifying the hydrothermal alteration zones, mapping lineament density, and revealing the dominant surface structures. The structural lineaments derived from the RS data and geological map have unveiled five main trends: NW–SE, NNW–SSE, N–S, NNE–SSW, and NE–SW. These trends likely play an essential role in the existence and distribution of radioactive anomalies, side by side with rock composition. Notably, the analysis of airborne magnetic data indicates that the NNW–SSE fault related to the Red Sea rift system is the predominant trend in the study area. Overall, this multifaceted approach provides insights into the radioactive mineralization and structural framework of the study area.

Location Intelligence in Sport Tourism Economics: Temporal Evolution and Sectoral Variation of Bassmaster Elite Series Fishing Tournaments

This study examines the economic and spatial impacts of Bassmaster Elite Series fishing tournaments on host communities through a novel application of location intelligence data. Using Advan’s monthly patterns dataset from 2022–2024, visitation patterns, consumer behavior, and spatial dependencies are analyzed across multiple tournament locations in South Carolina. Our spatially aware difference-in-differences framework incorporates treatment intensity based on proximity to tournament venues and explicitly accounts for spatial autocorrelation effects. Results reveal significant increases in business activity near location of tournament sites, with spatial clustering (Moran’s I = 0.333, p < 0.001) that strengthens over time (2022: 0.122, 2023: 0.171, 2024: 0.203). Effects persist beyond the immediate tournament period, with statistically significant impacts observed up to three months post-event, particularly for hospitality, recreation, and retail sectors. Robustness checks using alternative distance specifications and placebo tests confirm the validity of our findings. This research advances the understanding of how professional fishing tournaments generate economic value through both direct spending and spatial spillovers, providing valuable insights for event organizers, destination marketers, and local policymakers. Our methodological approach demonstrates the utility of geo-behavioral data in quantifying the community-level impacts of sporting events beyond traditional economic impact assessments.

The impact of population migration on public educational resource allocation between urban and rural areas

The allocation of public educational resources between urban and rural areas is one of the major topics in education research and is largely influenced by population shifts. Therefore, this study systematically reviews the burdens on urban educational resources, the impact on rural education, and the effects of two-way migration between urban and rural areas and illustrates the impact of population migration on urban-rural public educational resource allocation. To address these problems, this study proposes an optimization path that links population and resources with the utilization of big data. It also points out that it is necessary to promote the coordinated urban-rural development strategy, advance school consortium models, and drive the progress of rural schools by leveraging urban school resources. Government departments should increase fiscal transfer payments, gradually deepen the household registration reform, and ease restrictions on migrant children taking high school and college entrance exams in their places of residence. Aimed to promote education equality and improve education quality and effectiveness, this study proposes relevant suggestions.

Multiple Unmanned Aerial Vehicle Formation Control through Deep Reinforcement Learning with Offline Sample Correction

Multiple unmanned aerial vehicle (multi‐UAV) formation plays a pivotal role in both military and civilian applications. However, enabling multi‐UAV to efficiently acquire collaborative formation policies is challenging, primarily due to the complex kinematics and dynamic environmental factors involved. Herein, a deep reinforcement learning‐based multi‐UAV formation control approach is proposed. The multi‐UAV formation control problem as a Markov decision process to capture the inherent sequential decision‐making process is first modeled. Furthermore, an enhanced actor‐critic algorithm called offline sample correction actor‐critic (OSCAC) to learn the formation control policy is proposed. The core insight behind OSCAC is to optimize the utilization of historical data through correcting offline samples. Specifically, OSCAC adjusts the importance weights of the samples based on the discrepancies between the current policy and previous policies during each update step, making better use of past experience and improving the learning performance. The effectiveness and reliability of the proposed approach are validated through numerical simulations, software‐in‐the‐loop simulations, and real‐world experiments.

Determinants of Health Data Utilization by Posyandu Cadres for Toddlers as a Stunting Prevention Effort in Geneng Subdistrict, Ngawi Regency

Stunting remains one of the major public health issues at the national level in Indonesia. As an archipelagic country, Indonesia faces unique challenges in tackling stunting, particularly in regions with limited access to healthcare services. Advances in information technology offer new opportunities to support stunting prevention efforts, including through the utilization of health data. Such data can be used to detect stunting risks early and to monitor children's nutritional status more effectively. The use of health data applications or systems by Posyandu cadres is influenced by various factors, including availability of time, cost, level of trust, and perceptions of ease of use and usefulness. This study aims to identify the determinants of health data utilization by Posyandu cadres for toddlers as a stunting prevention effort in Geneng Subdistrict, Ngawi Regency. This research is an analytical quantitative study with a cross-sectional approach. A sample of 80 Posyandu cadres for toddlers in Geneng Subdistrict was selected using purposive sampling. Data were collected through questionnaires and analyzed using univariate, bivariate, and multivariate logistic regression tests to identify the factors influencing the use of health data in stunting prevention. The results show that the significant determinants include the age of the Posyandu cadre, their education level, and the amount of time they dedicate to Posyandu activities. The determinants of age, education level, and time significantly influence the utilization of health data and thereby affect the optimization of stunting prevention. Strengthening cadre capacity in these aspects is necessary to support more effective stunting prevention.

Smart weather: Why does it matter for urban studies?

ABSTRACT Smart weather services involve the collection, analysis and utilization of weather data through technology, and are now reshaping how cities govern their atmospheres. Operating through sensors, digital platforms, and municipal infrastructures, these systems promise hyperlocal, real-time responses to weather volatility. This intervention argues that smart weather constitutes a new mode of urban climate governance, embedding the atmosphere within regimes of prediction, modulation, and anticipatory control. Urban meteorology is being reconfigured through citizen sensing, proprietary platforms, and predictive services. By situating smart weather at the intersection of smart urbanism, climate adaptation, and urban meteorology, the paper calls for a critical urban geography of atmospheric governance focused on questions of spatial justice, infrastructural politics, and the right to climate-resilient urban life.

Presenting summary reports to patients with chronic health conditions to support values-based discussions and decision-making: A literature review.

Presenting summary reports to patients with chronic health conditions to support values-based discussions and decision-making: A literature review.

A robust single-antenna GNSS/MEMS fusion structure for reliable attitude determination in challenged environments

This study proposes an innovative single-antenna GNSS/MEMS integrated attitude determination system designed to address the challenges of low accuracy and poor robustness in urban complex environments, where conventional multi-antenna GNSS systems and standalone MEMS sensors often fail due to signal obstructions and drift errors. The key innovations include: (1) a hybrid TDCP/TDPR (Time Difference Carrier Phase/Pseudo-Range) technique to enhance GNSS-derived pitch and heading estimation, significantly improving data utilization; (2) a multi-strategy quality control framework incorporating Zero Velocity Detection (ZVD), Magnetic Disturbance Detection (MDD), and Fault Detection and Exclusion (FDE) to suppress outlier contamination and ensure measurement reliability; and (3) an adaptive noise parameter tuning mechanism to optimize filter performance. Experimental results demonstrate that the proposed system outperforms conventional SPP/MEMS, standalone MEMS, and baseline GNSS/MEMS fusion methods, achieving CDF95 attitude errors of 1.2785° (roll), 0.9412° (pitch), and 1.1991° (yaw)—the highest accuracy and robustness reported for single-antenna GNSS/MEMS systems in urban scenarios.

Construction and application of knowledge graph in digital humanities: a case study of Qiaokan

Abstract This study utilizes Knowledge Graph (KG) technology, taking Qiaoxiang Memory Heritage “Gu Zhai Yue Kan” and “Guang Yu Yue Bao” as examples, to analyze the contributions of overseas Chinese to the development of their hometown. By integrating multisource heterogeneous data, constructing datasets, and mining deep information, the practicality and efficiency of KG in managing Qiaoxiang Memory Heritage (Qiaokan) data are verified. KG not only provides data analysis capabilities but also offers precise data management tools for governments and research institutions through knowledge question answering and visualization technology. The research finds that overseas Chinese have made continuous investments in multiple fields, having a profound impact on their hometown’s development. KG demonstrates applications in information querying, knowledge reasoning, and discovery, capable of rapidly capturing donations and changes. The conclusion points out that KG is a preferred method for analyzing donation information, though it has limitations such as a limited sample size. Future improvements are needed in areas like automatic entity extraction and multimodal data mining. This study provides a new pathway for the storage, management, and utilization of cultural data in Overseas Chinese hometowns and offers technical support and a knowledge framework for scientific research.

The 6C Framework to Build a Connected Factory

Abstract Embracing Industry 4.0 (I4.0) relies on the crucial role of data collection and utilization. The benefits of I4.0 drive operational excellence in connected factories by elevating productivity, uptime, and quality. While the rationale of Industry 4.0 adoption is widely acknowledged, the challenge lies in the practical implementation of a connected factory. The resolution of both technical and human challenges is required for successful adoption. The proposed 6C Framework is structured around six key components to solve these challenges: Criteria, Connect, Communicate, Collect, Consume, and Culture. Understanding this high-level framework is foundational, as it guides countless strategic management decisions when implementing data collection. This framework is deconstructed into tactical aspects to ensure proper technical and cultural questions are considered throughout the data collection process. The 6C Framework addresses the human and technical aspects of data collection to aid in implementing I4.0.

Utilization of Remote Sensing Data in Determining the Threshold Value of Urban Ecological Quality Index in Bandung City, West Java, Indonesia

Bandung City has the highest land conversion rate in Indonesia and was named a city with a moderate environmental quality index status in 2022. This status has been exacerbated by the diminishing green spaces in the city due to rapid urbanization. Conducting ecological assessments has become increasingly important, one approach being the utilization of remote sensing data. Remote sensing data, specifically Landsat 8 OLI/TIRS, processed to derive the RSEI (Remote Sensing Ecological Index) based on the PCA value of PC1, requires further development. Several limitations of the RSEI in assessing ecological quality, such as the subjectivity of remote sensing data, the use of equal interval methods for index classification, and the inability to validate the results, are the focus of development in this study. Based on these weaknesses, the RSEIT offers advancements in integrating actual data to support RSEI, determining index thresholds, and enabling model validation. The findings of this study demonstrate that: (1) ecological issues such as floods, waste accumulation, and landslides are the most prevalent problems in the study area; (2) compared to RSEI, which relies solely on remote sensing data, RSEIT is a model that can be validated with actual data. During the dry and rainy seasons, it achieves threshold values of 0.474 and 0.566, respectively, demonstrating a model performance accuracy exceeding 70%. The average validation results show an overall accuracy of 83.34%, a sensitivity of 78.55%, and a specificity of 87.50% across both seasons; and (3) urban centers, characterized by extensive surface hardening, minimal vegetation, and numerous ecological issues, predominantly fall under the poor RSEIT category, especially during the dry season. In contrast, suburban areas with higher proportions of green space and fewer ecological problems are largely classified under the good RSEIT category, particularly during the rainy season. This study can be further enhanced by refining the threshold aspects and strengthening actual data collection through the involvement of various stakeholders with expertise in ecology.

Species Distribution Modeling of Killer Whales (Orcinus orca) in Australian Waters

ABSTRACTThe killer whale (Orcinus orca) is a globally distributed apex predator. This species is represented by distinct ecotypes or forms, which are well documented in the Northern Hemisphere and Antarctica. However, less is known about killer whales in Australia. While research efforts have been made to study these animals, a broader understanding of their range and drivers of occurrence is lacking. In this study, we model the spatial distribution of killer whales in Australian waters to identify potential areas of habitat suitability and conservation priority. A total of 1310 sightings were compiled, of which 1115 were used alongside a suite of static and dynamic predictor variables to build presence‐only MaxEnt species distribution models (SDMs) in three separate study areas: southeast (SE), southwest (SW) and northwest (NW) Australia. The SDMs identified potential areas of habitat suitability both within and outside of known locations for killer whales in Australia. All three models returned good discriminative power between presence and background points. However, good predictive power was only suggested for the SE and NW. The importance of certain predictor variables indicated a preference for different environmental conditions, supporting the notion of at least two ecologically distinct groups. Killer whales frequenting the SE and SW preferred temperate waters, whereas those in the NW preferred tropical waters. This work greatly increases our understanding of killer whales in Australian waters and identifies potential areas of biological importance for management and monitoring. It also complements ongoing research into their genetics, feeding ecology, and diversification, showcases the utility of citizen science data, and informs the conservation of this species, which is still considered data deficient and remains to be adequately protected under Australian Government legislation.

Health professional’s data management practice in Ethiopia: a systematic review and meta-analysis

BackgroundA global focus is producing high-quality data for use in decision-making at all levels of the health system. Healthcare professionals’ data handling techniques have been the subject of numerous research, yet the findings have been wildly inconsistent in Ethiopia. This inconsistency has not been thoroughly examined, leaving a gap in reliable evidence regarding the overall magnitude of effective data management practices within the healthcare system of Ethiopia.ObjectiveThis review aimed to determine the pooled magnitude of data management practices amid healthcare professionals in Ethiopia.MethodsWe developed a search strategy and conducted a comprehensive search across PubMed, Web of Science, African journals of online, Ethiopian journal of health science, Scopus, Medline, Embase, WHO Global Health Library, Google Scholar, Scopus, HINARI, CINAHL, and the Cochrane Library to identify relevant studies between October 10 and November 10, 2024. The Joanna Briggs Institute quality assessment methodology, which has nine categories, was used to evaluate each study’s quality. Microsoft Excel version 2019 and Stata software was used to extract and analyzes the data. We calculated total heterogeneity across studies using the I2 statistic and estimated the pooled rate of good data management practices with a fixed-effects model at a 95% confidence interval.ResultsIn the first phase, 943 publications on data management practices of healthcare professionals were identified from multiple databases between October 10 and November 10, 2024. A total of 221 studies were excluded due to duplication, and an additional 705 studies were removed for being irrelevant to the outcome of interest. An additional 10 articles were excluded due to missing outcome data or unavailable full texts. Ultimately, 7 studies involving a total of 2654 health professionals were included in the meta-analysis. The overall pooled estimate of good data management practice in Ethiopia was 64.4% (95% CI, 52.27–76.47%). The subgroup analysis revealed the highest proportion in Gondar at 73.02% (95% CI, 52.33–93.70%), and the lowest proportion was observed in the Oromia Special Zone, at 53.62% (95% CI, 29.21–78.02%). Similarly, studies conducted before 2019 reported a slightly lower prevalence (63.60%; 95% CI, 44.76–82.45%) compared to those conducted after 2019 (64.91%; 95% CI, 49.13–80.69%). However, the subgroup analysis indicated that these differences were not statistically significant across study years or regions.ConclusionThe pooled meta-analysis result revealed that approximately six out of ten healthcare professionals in Ethiopia engage in good data management practices, highlighting a significant improvement in this critical area. Although there were a geographical variations and study year, it was not statistically significant. These results underscore the need for targeted interventions to enhance data management practices. By investing in training, improving infrastructure, and fostering a culture that values data utilization, stakeholders can significantly elevate the standard of data management among healthcare professionals. Addressing these gaps is essential for strengthening health information systems and ultimately improving health outcomes across Ethiopia.

Open Science, Data, and Methodologies: Lessons learned from the NIHR-RESPIRE Network in Asia

NIHR-RESPIRE, a Global Health Research Unit funded by NIHR, is committed to advancing respiratory health research in Asia. We prioritise Open Science, Data, and Methodologies to maximise research data utility securely, sharing the lessons encountered across seven LMIC partner countries (Bangladesh, Bhutan, India, Indonesia, Malaysia, Pakistan, and Sri Lanka). Our strategic shift from traditional data sharing to LMIC-tailored Open Science practices ensures data privacy and security. This includes refining Data Management Plans, metadata standards, and mandating FAIR Data sharing, providing methodological support, and developing Open Science Policy Guidelines. We advocate for the adoption of open science principles to maximise secure data use and value with a focus on FAIR data. We also provide aid to partners in enhancing their data-related skills, hosting regular meetings, and establishing internal data monitoring structures to bolster cross-cutting activities within RESPIRE. Through capacity building, we have enabled high-quality respiratory health research using Open Science principles, enhancing data sharing efficiency, research visibility, and ultimately respiratory health outcomes in Asia and beyond. Our experience underscores the following lessons: Flexibility in data sharing, tailored to LMIC researchers' needs, is essential; Training and support to enhance knowledge of methodologies and dispel misconceptions are key to successful data stewardship; Appointing a focal person for structured anonymised data sharing and supporting the internal Data Monitoring Committee are critical. We recognise Open Science's potential to foster innovation, collaboration, and knowledge sharing in respiratory health research.

Contributions of District Health Information Software 2 (DHIS2) to maternal and child health service performance in Ethiopia: an interrupted time series mixed-methods study

BackgroundThe District Health Information Software 2 (DHIS2) is the primary digital platform for health management information systems (HMIS) in Ethiopia, aligning with the nation’s digitization strategy. Despite widespread implementation, its effectiveness on key health service indicators, particularly maternal and child health (MCH) services, remain unclear.ObjectiveThis study aimed to evaluate the contribution of DHIS2 on data use and the performance of selected MCH indicators in Ethiopia, comparing data before and after the implementation of DHIS2.MethodsWe analysed data from primary health care units (PHUs) across five diverse regions of Ethiopia, encompassing urban, agrarian, and pastoralist settings. A mixed-methods approach was employed, combining quantitative and qualitative methods to provide a comprehensive understanding of the data. The quantitative component involved examining performance reports of selected maternal and child health (MCH) indicators from 2013 to 2022 to assess changes before and after the implementation of the District Health Information System 2 (DHIS2). Data were collected electronically and analysed using descriptive statistics and interrupted time series (ITS) analyses to identify trends and patterns. The qualitative component included interviews and focus group discussions with health workers and stakeholders to explore contextual factors influencing MCH service utilization and performance. The data were analysed thematically using OpenCode 4.1 software.ResultsThe implementation of DHIS2 significantly contributed to the enhancement of MCH data utilization within PHU facilities. This improvement supported decision-making processes in various aspects of maternal and child healthcare delivery, including target setting, resource allocation, program monitoring, and clinical service provision. Specifically, DHIS2 led to increased monthly mean performance of key indicators such as antenatal care visits, skilled birth attendance, and immunization rates. Notable improvements in service delivery were observed, with significant increases in institutional delivery rates over time.ConclusionsThe study highlights DHIS2’s significant contribution to improving MCH services in Ethiopia, with increased institutional delivery rates and ANC coverage reflecting enhanced data-driven decision-making. Most facilities relied on DHIS2 for resource allocation and program monitoring, though challenges like offline usage and accessibility persist. To maximize impact, improving offline data management, training staff, leveraging real-time reporting, and addressing accessibility through connectivity investments are recommended.

Unitanony: a fine-grained and practical anonymization framework for better data utility

In order to share data without revealing private information, privacy-preserving data publishing techniques are proposed. K-anonymity and l-diversity secure against identity and attribute disclosure. Anonymization algorithms enforce the above models and are willing to reach two primary goals: achieving the privacy objective while maximizing data utility. Even though anonymization has been studied for decades, finding efficient techniques to improve data utility is an open question. It is a crucial challenge that impacts many anonymized data on the web, cloud, and IoT environments. However, some factors incur huge information loss for existing works. The original dataset may be transformed into generalized data to an excessive extent. To address this problem, we give a new framework and propose a heuristic algorithm called UnitAnony. It builds a full-coverage hierarchy for more generalization candidates and proposes an interval-mapping technique for fine-grained generalization extent. However, these improvements raise another challenge. It is the vast cost because more generalization will derive many operations for forming new values, grouping records, and verifying anonymization models. Therefore, we designed a data structure unit to generalize records at low costs and implemented a skipping strategy to execute the algorithm within an acceptable time. Besides, our algorithm can support many models. By evaluating well-known K-anonymity and l-diversity on real-world datasets, i.e., Adult and Census datasets, the experimental results demonstrate that our algorithm outperforms the existing algorithms (e.g., Mondrian, Top-down, Improved-Clustering, Flash, and Incognito) regarding data utility and effectiveness.

Improving the usability of large emergency 911 data reporting systems: A machine learning case study using emergency incident descriptions.

Improving the usability of large emergency 911 data reporting systems: A machine learning case study using emergency incident descriptions.

Locally Differentially Private Frequency Estimation via Joint Randomized Response

Local Differential Privacy (LDP) has been widely recognized as a powerful tool for providing a strong theoretical guarantee of data privacy to data contributors against an untrusted data collector. Under a typical LDP scheme, each data contributor independently randomly perturbs their data before submitting them to the data collector, which in turn infers valuable statistics about the original data from received perturbed data. Common to existing LDP mechanisms is an inherent trade-off between the level of privacy protection and data utility in the sense that strong data privacy often comes at the cost of reduced data utility. Frequency estimation based on Randomized Response (RR) is a fundamental building block of many LDP mechanisms. In this paper, we propose a novel Joint Randomized Response (JRR) mechanism based on correlated data perturbations to achieve locally differentially private frequency estimation. JRR divides data contributors into disjoint groups of two members and lets those in the same group jointly perturb their binary data to improve frequency-estimation accuracy and achieve the same level of data privacy by hiding the group membership information in contrast to the classical RR mechanism. Theoretical analysis and detailed simulation studies using both real and synthetic datasets show that JRR achieves the same level of data privacy as the classical RR mechanism while improving the frequency-estimation accuracy in the overwhelming majority of the cases by up to two orders of magnitude.

The candidate gene OLFML2A possibly contributed to the variation of total number of teats in Meishan and Erhualian pigs by influencing the formation of mammary placodes.

The candidate gene OLFML2A possibly contributed to the variation of total number of teats in Meishan and Erhualian pigs by influencing the formation of mammary placodes.