Data-driven Analysis Research Articles

Data-Driven Prediction of Reactivity and Additive Selection for C(sp2)-(Hetero)atom Bond Couplings in an Adaptive Dynamic Homogeneous Catalysis.

Under visible light-driven redox conditions, employing transition-metal catalysis provides a powerful platform for constructing C(sp²)-(hetero)atom bonds. Although these reactions are highly significant, they require precise optimization of reaction parameters. König, Ghosh and colleagues introduced an adaptive dynamic homogeneous catalysis (AD-HoC) platform that furnishes robust, high-yield conditions for photocatalyzed cross-coupling reactions. The AD-HoC system eliminates the need to optimize catalysts, ligands, and bases; instead, it achieves C(sp²)-(hetero)atom bond coupling by merely altering additives and substrate molecules. Leveraging the predictability of reaction conditions within the AD-HoC system, machine learning offers a method to evaluate the reactivity of substrate combinations and the categories of additives. Our research integrates high-throughput quantum mechanical calculations with cheminformatics approaches to explore the reactivity of substrate combinations and the selection of additives within the AD-HoC system. Further data-driven analysis reveals that the electronic characteristics of electrophiles and the geometric characteristics of nucleophiles are key factors regulating reactivity within the AD-HoC system. Herein, we present an end-to-end tool for prediction starting from the SMILES representation. This work demonstrates the collaborative use of computational statistics and machine learning to predict the reactivity and reaction conditions of substrate combinations, thereby enhancing the precision and efficiency of synthetic processes.

Abstract 3714: Microscopic mayhem: Serial killing and other immune cell dynamics captured in situ

Critical to advancing immunotherapy and cell therapy in cancer is developing a deeper understanding pf the dynamics of immune cell-mediated cytotoxicity. The results from the multidisciplinary effort reported here include numerous movies of immune cell-mediated cytotoxicity with striking examples of serial killing, foraging, path-tracking, cytokine gradients at tumor margins, and killing dynamics, in some cases revealing peak apoptotic signatures just minutes after T Cell engagement. In vitro studies of immune cell killing are traditionally performed using time-lapse imaging and biochemical assays, but these methods are often limited by spatial and temporal resolution, throughput, and the ability to extract the dynamics of cellular interactions. This study integrates high-resolution and high-speed laser scanning confocal microscopy with artificial intelligence (AI), and machine learning (ML) approaches to provide a high-resolution data-driven analysis of immune cell killing dynamics in vitro. We have engineered a perfusion-enabled 3D culture system integrated microscopy to assess cellular dynamics for extended periods of time. Perfusion culture maintains the interstitial flow of liquid culture media, clearing the microenvironment of toxic metabolites and reactive oxygen species. This platform uses a Liquid-Like Solids (LLS) to mimic the transport dynamics of a capillary bed. Integrated microscopy allows in situ quantification of spatiotemporal cytokine concentrations, immune cell tracking, immune cell killing dynamics, and invasion dynamics. In one example, we used Human CAR T cells with an anti-cd19.28z CAR that included a fluorescent mCherry protein fused to the carboxy terminus of the CAR vs. target cancer cells (Burkitt lymphoma) that express GFP. The media contained a 0.1% concentration of Propidium Iodide as a terminal maker of cell death. Such multi-channel data highlights the critical stages of immune-mediated cytotoxicity; however, the complexity and volume of confocal imaging data necessitates advanced computational tools for analysis. The ML algorithm used in this study comes from computer vision using object detection, which is a cornerstone technique in medical imaging. We utilized the YOLOv8X model, which is distinguished by its high accuracy and efficiency in object detection tasks. The YOLO (You Only Look Once) architecture is designed to predict both bounding boxes and class probabilities in a single forward pass, making it exceptionally fast and suitable for real-time applications. Finally, by integrating robotics, imaging, and AI/ML detection we are able to capture immune cells of interest in situ. This integrated analysis of complex imaging data from confocal microscopy has been significantly accelerated, enabling the identification of subtle patterns and dynamic interactions between immune cells and their targets, and the actual collection of cells of interest. Citation Format: W. Gregory Sawyer, Diego Pedro, Alfonso Pepe, Reginald Atkins, Chandler Gatenbee, Luis Felipe, Carlos Garcia Fernandez, Soraya Zorro Manrique, Said Cifuentes Maury, Athena Tsingelis, Duy Nguyen, Alexander Anderson, Erin George, Gilmer Valdes, Fredrick Locke, Patrick Hwu. Microscopic mayhem: Serial killing and other immune cell dynamics captured in situ [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 3714.

Evaluating Competency Development and Academic Outcomes: Insights from Six Semesters of Data-Driven Analysis

Competency-Based Education (CBE) has been widely studied since the 1970s, yet it remains innovative due to its challenges across disciplines and cultures. Tecnológico de Monterrey, a Mexican private institution, implements CBE through its Tec21 model, which emphasizes challenge-based learning to develop disciplinary and transversal skills. Since its launch in 2019, Tec21 has generated extensive data, offering an opportunity to assess its performance and ensure quality. This study analyzes data from six academic semesters in the School of Engineering and Sciences to address key quality assurance questions. First, we evaluate whether initially enrolling in a generic area before selecting a specific program improves long-term student outcomes. Second, we examine competency development, identifying challenges in achieving certain skills and their links to dropout rates and module difficulty. Third, we explore the relationship between final grades, module credit allocation, and Tec weeks to assess curriculum alignment with academic performance. Using data from over 550,000 evaluations of 4500+ students, our analysis provides robust quality metrics. Findings suggest that students who start in generic areas perform better long term and highlight the complex interplay between competencies, module characteristics, and academic success. These insights deepen the understanding of CBE implementation and offer recommendations to improve educational strategies and quality assurance within competency-based frameworks.

A Structured Review of the Effects of Social Media Marketing on Consumer Behavior.

Social Media has transformed the marketing landscape, reshaping how businesses interact with consumers and influence purchasing decisions. This study examines the role of social media marketing (SMM) in shaping consumer behavior, emphasizing its impact on product discovery, brand interaction, and decision-making. A data-driven analysis revealed a moderate positive correlation (r = 0.629) between social media marketing and consumer purchasing behavior, highlighting the potential of SMM to influence consumer perceptions and foster trust.Social media platforms, such as Facebook, Instagram, and TikTok, allow businesses to engage directly with consumers through personalized content, influencer collaborations, and user-generated reviews. These interactions play a significant role in both pre-purchase and post-purchase phases, helping consumers form brand attitudes, make informed decisions, and share their experiences with others. This shift from traditional mass media to digital platforms has empowered consumers and strengthened brand-consumer relationships.While the benefits of SMM are substantial, including improved brand visibility and loyalty, challenges persist. Content saturation and the potential for negative feedback demand innovative and strategic approaches to stand out and maintain consumer trust. Businesses must remain agile, addressing consumer concerns transparently and adapting to evolving trends to ensure long-term success.In conclusion, social media marketing is a powerful tool for influencing consumer behavior, offering opportunities to build trust, foster loyalty, and create meaningful consumer connections. By navigating its challenges effectively, businesses can leverage SMM to achieve sustainable growth in an increasingly competitive digital marketplace. Keywords: Social Media Marketing, Consumer Behavior, Digital Marketing, Brand Perception, Influencers, Consumer Engagement

Understanding Unique Behavioral Patterns through Multimodal Analysis of Eye-Hand Coordination in Autistic Children (Student Abstract)

Data-driven analysis has shown promising results in identifying subtle patterns in the behavior of individuals with Autism Spectrum Disorder (ASD) for diagnosis and intervention. However, most existing methods primarily focus on a single behavioral modality (e.g., eye movements) instead of capturing the intricate multimodal behavior of humans. We propose a multimodal approach that investigates the underlying connections between eye movements and hand motions through eye-to-hand prediction. To tackle the highly noisy and irregular behavioral data, we propose a novel approach that defines the prediction as a machine translation problem and leverages a sequence-to-sequence machine learning model for the prediction. An experimental study on a dataset collected from a VR system has demonstrated high prediction accuracy. The significant difference in the prediction accuracy between the autistic group and their typically developing (TD) peers serves as quantitative evidence to objectively understand the restricted and repetitive behaviors (RRBs) in autistic children. The source code can be accessed here: https://github.com/mathjams/AAAI_2024.

Facilitators and barriers to establishing lactation among women with diabetes?

To identify the facilitators and barriers to establishing lactation among women with diabetes in pregnancy. A qualitative descriptive study involving face-to-face semi-structured interviews with 13 postpartum women. An inductive, data-driven content analysis approach was utilised. This study followed the consolidated criteria for reporting qualitative research (COREQ) guidelines. Facilitators to breastfeeding and establishing lactation among women with diabetes included antenatal colostrum harvesting during the antenatal period and support from health care professionals. Barriers included emotional self-efficacy challenges and a lack of support from healthcare professionals in the hospital setting. Women with diabetes report further challenges establishing breastfeeding due to delayed lactogenesis ll (secretory activation), hypoglycaemia and self-efficacy challenges. Given the adverse health risks for mother and infant associated with not breastfeeding, women with diabetes require individualised person-centred lactation care plans to maximise their opportunity to experience a successful breastfeeding journey.

Evaluation of therapeutic plasma exchange in anti-neutrophil cytoplasmic antibody-associated vasculitis in real-life settings: long-term results of propensity score matching analysis in high-risk patients

Objective To evaluate the prognostic significance and safety of therapeutic plasma exchange (TPE) in patients with anti-neutrophil cytoplasmic antibody-associated vasculitis (AAV) with severe organ or system involvement. Method Data of patients diagnosed with AAV between 2011 and 2021 were evaluated retrospectively. Patients with baseline estimated glomerular filtration rate (eGFR) ≤ 50 mL/min/1.73m2 or diffuse alveolar haemorrhage (DAH) were included in the analysis (n = 71). Patients who underwent TPE were compared with others in the groups formed after two models of propensity score matching (PSM). Initial PSM was performed according to the presence of DAH, age, gender, eGFR, and BVAS. A data-driven approach was conducted for the second model. Results In the initial PSM cohort (n = 48), the remission rate at 6 months was lower in the TPE group (p = 0.04). There were no significant differences in mortality and improvement of eGFR at 6 and 12 months. No severe complications due to TPE were observed. Rates of serious infections (SIs) and end-stage renal disease (ESRD) at 12 months were higher in the TPE group (p = 0.016 and 0.02, respectively). Data-driven PSM analysis (n = 44) revealed no significant differences between groups. Conclusion We did not demonstrate a positive effect of TPE on remission, ESRD, and mortality in AAV in this study. Despite low short-term complication rates, the increased risk of SIs possibly associated with ESRD was remarkable. The limited long-term benefits of TPE should be carefully weighed against its associated risks when selecting patients for treatment.

The spatial transcriptome of the late-stage embryonic and postnatal mouse brain reveals spatiotemporal molecular markers

The neocortical development process includes cell proliferation, differentiation, migration, and maturation, supported by precise genetic regulation. To understand these processes at the cellular and molecular levels, it is necessary to characterize the fundamental anatomical structures by gene expression. However, markers established in the adult brain sometimes behave differently in the fetal brain, actively changing during development. The spatial transcriptome is a powerful analytical method that enables sequence analysis while retaining spatial information. However, a deeper understanding of these data requires computational estimation, including integration with single-cell transcriptome data and aggregation of spots at the single-cell cluster level. The application of such analysis to biomarker discovery has only begun recently, and its application to the developing fetal brain is largely unexplored. In this study, we performed a spatial transcriptome analysis of the developing mouse brain to investigate spatio-temporal regulation of gene expression during development. Using these data, we conducted an integrated study with publicly available mouse data sets. Our data-driven analysis identified novel molecular markers of the choroid plexus, piriform cortex, and thalamus. Furthermore, we identified a novel molecular marker that can determine the dorsal endopiriform nucleus (DEn) of the developmental stage in the claustrum/DEn complex.

Predicting Pulsed-Laser Deposition SrTiO3 Homoepitaxy Growth Dynamics Using High-Speed Reflection High-Energy Electron Diffraction.

Pulsed-laser deposition (PLD) is a powerful technique for growing complex oxides with controlled stoichiometry. To understand growth dynamics therein, it is common to leverage in situ spectroscopies, such as reflection high-energy electron diffraction (RHEED), to monitor surface crystallinity. Most commercial systems rely on video-rate cameras operating at 60-120 Hz that lack sufficient temporal resolution to capture growth dynamics at practical deposition frequencies. Here, a high-speed platform to record in situ dynamics via RHEED at >500 Hz is implemented. An open-source analysis package is designed to fit diffraction spots to 2D Gaussians, allowing single-pulse surface reconstruction kinetics extraction. Using homoepitaxially deposited (001)-oriented SrTiO3 as a model system, we demonstrate how high-speed RHEED can provide real-time insight into growth processes obscured by slower acquisition systems. By fitting the single-pulse intensity to a set of exponential functions, we observe changes in the characteristic decay time and mechanism correlated to the substrate step width and surface termination. We observe distinct surface effects, with diffraction intensity decaying on lower-energy TiO2-terminated surfaces and stabilizing on SrO- or mixed-terminated surfaces. Similarly, using an exponential model, the extracted characteristic time of adatom deposition decreases with increased density of bonding sites associated with mixed termination and narrower step widths. Ultimately, this work shows how increasing RHEED temporal resolution can uncover new insights into growth processes, with practical implications for the design and control of PLD processes. This experimental platform provides new capabilities to enable data-driven machine learning analysis and autonomous control systems to enhance the complexity and fecundity of PLD.

Building operator-centric digital twins for optical transport networks: from data models to real applications

The complexity and heterogeneity of modern optical transport networks (OTNs) demand advanced solutions to enhance their operation and maintenance. This paper presents lessons learned from the design and implementation of a digital twin network (DTN) tailored to network operators’ requirements, integrating hybrid models that combine physical representations with data-driven analysis. By leveraging telemetry data, the DTN can facilitate network operators to perform predictive maintenance, resource optimization, and risk analysis. Through a case study on soft failure detection across 454 optical sections, we analyze the value of transforming theoretical concepts coupled with collected data into actionable insights, reducing operational overhead and enhancing network resilience. This work underscores the potential of hybrid modeling in the context of DTN to create operator-centric solutions for optical transport networks.

Maternal-offspring brain and tissue cross-talk in preeclampsia: insights from a rat model

This study aimed to investigate the differential metabolic profiles across maternal and offspring brains, serum, and placental tissues in preeclampsia (PE), with a particular focus on elucidating the maternal-offspring brain and tissue cross-talk that may contribute to the complex pathophysiology of PE. PE was induced in rats using the nitric oxide synthase inhibitor N-nitro-L-arginine methyl ester (L-NAME) to simulate both early-onset PE (EOPE) and late-onset PE (LOPE). We utilized non-targeted proton nuclear magnetic resonance (NMR) metabolomics to characterize the metabolic profiles of serum, placental tissue extracts, and brain tissues from both mothers and offspring. Multivariate analysis, Spearman correlation, Density-Based Spatial Clustering of Applications with Noise algorithm, Data-Driven Statistical Predictive Correlation network analysis and Tissue heterogeneity analysis were employed to explore tissue-specific metabolic signatures and their interactions. Following L-NAME induction, both EOPE and LOPE presented significant metabolic differences and shared traits across tissues, with distinct tissue-specific responses characterizing the metabolic profile of PE. Serum from both PE groups showed a decrease in tryptophan, isobutyrate, and lactate, with an increase in betaine. Lactate was upregulated in placental tissues, highlighting its metabolic role. Extensive intra-tissue metabolic correlations and inter-tissue metabolite exchanges were detected among the maternal brain, serum, placenta, and offspring brain across all three experimental groups. EOPE and LOPE exhibited distinctly different metabolic characteristics and trajectories of differential metabolites, along with diverse interaction patterns between the maternal/offspring brain and the placenta. This study uncovers the multi-tissue metabolic remodeling in response to preeclampsia, implying that addressing pathophysiological stress is crucial and may have potential implications for neurological outcomes. The comprehensive analysis highlights the pivotal role of the brain-placenta axis in preeclampsia, advocating for a classified diagnostic and management approach.

An Investigation into the Information-based Teaching Ability of English Teachers in Chinese Vocational Colleges from the Perspective of TPACK Theory

Grounded in the TPACK framework, this study examines the current state and influencing factors of information-based teaching ability among 176 English teachers from seven vocational colleges in Anhui Province through a questionnaire survey. Findings reveal that while teachers demonstrate strong content knowledge (CK) and pedagogical knowledge (PK), their instructional flexibility and exploration of diverse teaching styles remain limited. Although initial alignment between technological tools and instructional goals has been achieved, the deeper role of technology in reshaping teaching paradigms is underutilized. A lack of motivation for technological advancement is evident, with teachers having ≥16 years of experience lagging in technological knowledge (TK). Additionally, a significant gap exists in data-driven student learning analysis within pedagogical content knowledge (PCK). The weakest aspect is the reflective integration of TPACK, with a general trend of low-level homogeneity, indicating insufficient teacher collaboration and technology-sharing awareness. Notably, the enhancement of TPK significantly contributes to TPACK development, while self-efficacy emerges as a critical determinant. Based on these findings, the study proposes strategies to enhance information-based teaching ability, aiming to foster comprehensive improvement among vocational college English teachers.

Study On Student Centric Quick Commerce Application

Quick commerce (q-commerce) has rapidly emerged as a transfo rmative force in the e-commerce sector, catering to the demand for ultra-fast delivery through hyperlocal logistics, AI-driven inventory management, and dark store networks. While existing q-commerce models have primarily targeted urban consumers, this research investigates its applicability and potential within student communities, a demographic with unique purchasing behaviors and immediate consumption needs. This study presents a data-driven analysis of market demand, operational feasibility, and financial sustainability for a student-centric q-commerce platform. Through primary research involving extensive student surveys and secondary data evaluation, the study assesses consumer behavior patterns, preferred product categories, and logistical challenges. The findings highlight the viability of a specialized model tailored to student communities, underpinned by AI-powered operational efficiencies. Additionally, the study explores historical investment trends in the q-commerce space, providing insights into the funding potential for such ventures. By establishing an academic foundation through empirical research and industry analysis, this paper offers a structured framework for evaluating q-commerce's role in student consumer ecosystems. The insights derived serve as a reference for future business models, investment considerations, and strategic decision-making in this evolving sector.

AlzAware: A Comprehensive Online Resource for Alzheimer’s Awareness and Support

Alzheimer's disease (AD) is a progressive neurodegenerative disorder that significantly impacts cognitive functions, often leading to memory loss and behavioral changes. Early detection is crucial for effective intervention and management. AlzAware, a comprehensive online platform, offers an innovative approach for the detection of Alzheimer's disease using cutting-edge diagnostic technologies such as Functional Magnetic Resonance Imaging (fMRI), Electroencephalogram (EEG), and advanced neurological diagnostics. Through the integration of pattern recognition algorithms and data-driven analysis, AlzAware provides an accurate and accessible tool for detecting AD at its earliest stages. The platform utilizes fMRI to visualize brain activity, EEG to analyze electrical patterns in neural activity, and sophisticated machine learning techniques to identify early signs of neurodegenerative conditions. By offering a non-invasive, user-friendly solution, AlzAware aims to improve the accuracy of Alzheimer’s diagnosis and provide valuable support for both clinicians and individuals at risk of developing Alzheimer’s disease. This paper discusses the underlying technologies behind AlzAware, its potential for enhancing diagnostic precision, and its role in advancing the fight against Alzheimer's disease. Keywords - fMRI (Functional Magnetic Resonance), Imaging, EEG (Electroencephalogram), Neurological Diagnostics, Neurodegenerative Conditions, Pattern Recognition.

The Integration of the RBL-STEM Learning Model and Graph Theory in Solving Transportation and Logistics Optimization Problems to Enhance Students' Computational Thinking Skills

Abstract This study aims to integrate the Research-Based Learning (RBL) model with the STEM approach and graph theory in solving transportation and logistics optimization problems to enhance students’ computational thinking skills. The model is designed to train students in identifying transportation and logistics issues, exploring solutions based on graph theory, and analyzing and optimizing distribution routes using algorithms such as Dijkstra and Floyd-Warshall. This research employs a narrative qualitative method, analyzing research-based learning through six RBL-STEM phases: problem identification, initial solution exploration, data collection, data analysis, result interpretation, and model presentation. The results show that this approach significantly improves students’ skills in problem decomposition, pattern recognition, algorithmic thinking, and abstraction in solving real-world problems. Concretely, students demonstrated an average score improvement of 23% in computational thinking indicators after participating in the learning process, particularly in designing optimal route graph models and constructing algorithmic logic for route selection. Moreover, the use of technologies such as Google Maps and graph modeling software enables students to perform real-time data-driven analysis. Therefore, this study proves that the integration of RBL-STEM and graph theory not only enhances conceptual understanding of transportation and logistics optimization but also equips students with applicable computational thinking skills for real-world contexts. This learning model can serve as a reference in developing more innovative and industry-relevant STEM-based teaching strategies. Keywords: RBL-STEM; Graph; Transportation; Logistics; Computational Thinking Skill.

Mapping the Determinants of Preclinical Medical Skills: A Data-Driven Analysis Enhanced by Vosviewer

A clear understanding of the factors affecting the development of preclinical medical skills plays a pivotal role in enhancing the quality of training and ensuring the clinical competence of the future healthcare workforce. Recent studies emphasize that the curriculum, resources, and teaching methods do not merely have isolated impacts but also exhibit a complex interaction. Evidence-based curricula, combined with modern simulation technologies and a highly specialized teaching staff, are regarded as essential solutions for improving skills and clinical problem-solving abilities. These strategies contribute to the creation of a high-quality healthcare workforce, capable of meeting the increasingly stringent demands of clinical practice (Sören Eric Huwendiek, 2016). The primary objective of this study is to identify and categorize the key determinants influencing preclinical medical skills through bibliometric analysis, combined with advanced statistical methods. Using VOSviewer, we visualize the interrelationships among these factors, providing a clear framework for reference. This research aims to highlight previously underexplored elements and offer actionable insights for optimizing preclinical medical education programs, thereby enhancing teaching methodologies and improving educational outcomes. This study employs a cross-sectional bibliometric analysis, integrating data from Scopus to analyze co-citation and co-authorship patterns to identify key factors influencing preclinical medical skills. Data were retrieved using the search formula: 'TITLE-ABS-KEY (Medical) AND (clinical) AND (skill) AND (factor*) AND ("affecting" OR "influencing").' The data set consists of 697 publications from Scopus, published between 2000 and 2025, focusing on studies related to medical education and preclinical skills. The VOSviewer tool (version 1.6.20) was selected for its ability to effectively visualize and map co-citation and co-authorship networks, providing clear insights into the relationships among influencing factors. Advanced statistical techniques, such as cluster analysis and network centrality measures, were used to identify the most influential factors and their interconnections. Metrics such as network density and time-slicing techniques were applied to analyze temporal changes in the research focus. This methodology has strategic significance in optimizing teaching strategies, thereby enhancing the quality of preclinical medical education. This study identifies key publishing trends in the medical field, showing significant growth since 2000, with a breakthrough in 2019. The United States leads in the number of publications, followed by the United Kingdom. Network analysis of collaborations highlights the strong partnership between the U.S. and the U.K., with frequent co-authorship links. Prominent keywords, including 'human,' 'medical education,' 'clinical competence,' and 'clinical practice,' reflect core themes in medical education, with 'clinical competence' showing a marked increase in publications after 2019. Leading journals, such as BMJ, JAMA, Acad Med, and Med Educ, exhibit substantial influence, with BMJ publishing over 35 articles on medical education between 2000 and 2025 on this topic. These journals significantly contribute to shaping the field's development. These findings provide strategic insights for guiding future research and improving medical education strategies, aimed at enhancing training quality and practical application. This study provides a comprehensive overview of the key determinants in the development of preclinical medical skills, highlighting the roles of teaching methods, technology integration, and student engagement. The findings hold strategic significance for shaping education policies based on scientific evidence and developing targeted teaching strategies tailored to practical needs, contributing to the enhancement of training quality. This research maps knowledge to support evidence-based decision-making, guiding the implementation of medical education activities and fostering further research, thereby improving preclinical medical skills training strategies.

Hybrid Data Driven Deep Learning Framework for Material Property Prediction

The research presents a hybrid approach of regression modeling with data-driven analysis for predicting steel's mechanical properties by analyzing the effects of composition on strength. The study fills the gap of models in accurately predicting steel's performance based on composition since traditional methods cannot fully capture complex relationships between alloying elements and material properties. Various regression models have been used for predicting material properties, such as Linear Regression, Random Forest Regression, Support Vector Regression (SVR), XGBoost Regression, and Neural Networks, and in this paper, Graph Attention Transformer Network (GAT-TransNet) is proposed. Incorporating novel graph attention into the transformer architecture model, GAT-TransNet handles complex data relationships and improves predictive accuracy. Data-driven analyses are also carried out alongside regression analysis to establish how alloying elements, such as carbon (C), manganese (Mn), and chromium (Cr), affect steel's mechanical properties strength, yield strength, hardness, and ductility. The study established that the GAT-TransNet model outperformed other regression models, with an R² score of 0.95, the lowest MAE of 1.40, and an MSE of 4.41, thus underscoring its superior predictive capability compared to existing models. Data-driven insights show that manganese hardens and increases wear resistance, while chromium enhances corrosion resistance and increases tensile strength. This has great importance for optimizing specific steel compositions for industrial applications. Combining machine learning methodologies with composition analysis, this study complements predictive modeling for steel properties with material design and promises better efficiency and targeting in steel production.

Big Data-Driven 3D Visualization Analysis System for Promoting Regional-Scale Digital Geological Exploration

As geological exploration technology advances, geoscience relies on digitization and intelligence to address challenges such as data fragmentation, multi-source heterogeneity, and visual analysis. This study develops a big data-driven 3D visual analysis system for regional-scale applications. The system integrates three core technological components: (1) a heterogeneous cloud resource scheduling method employing an optimized CMMN algorithm with unified cloud API standardization to enhance task distribution efficiency; (2) a block model-based dynamic data aggregation approach utilizing semantic unification and attribute mapping for multi-source geological data integration; (3) a GPU-accelerated rendering framework implementing occlusion culling and batch processing to optimize 3D visualization performance. Experimental validation shows the improved CMMN algorithm reduces cloud task completion time by 2.37% while increasing resource utilization by 0.652% compared with conventional methods. The dynamic data model integrates 12 geological data types across eight categories through semantic mapping. Rendering optimizations achieve a 93.7% memory reduction and 60.6% faster visualization compared with baseline approaches. This system provides robust decision support and reliable tools for the digital transformation of geoscience work.

Data-driven analysis and visualization of dielectric properties curated from scientific literature

Data-driven analysis and visualization of dielectric properties curated from scientific literature

Sustainable Co-Working Spaces in Education Institutions: A Data-Driven Analysis of Attributes Influencing Human Interaction

Co-working spaces have gained significant attention as flexible work environments. However, the specific attributes of sustainable co-working spaces that most influence human interaction, particularly in educational institutions, remain underexplored. There is a lack of such space in education institutions for students’ self-studying and self-development outside of formal learning hours. This study focuses on respondents within educational institutions, addressing a significant research gap in sustainable co-working space design for these settings. It aims to identify and evaluate the impact of sustainable co-working space attributes on human interaction. An online survey was developed on Google Forms and distributed via social media platforms, resulting in 65 responses from persons from various levels of education institutions. Using survey data collected from participants, the perceived impact of various attributes – ranging from functional attributes to sustainable attributes – was analysed. A combination of descriptive statistics, ANOVA tests, and post-hoc analyses were conducted to examine significant differences in the perceived impacts of these attributes. Results reveal that core attributes like internet access and open layouts significantly enhance human interactions, while sustainability-focused attributes like green furniture and energy-efficient designs exhibit lower perceived impact. The findings were translated into a prototype sustainable co-working space for a future validation study. These findings underscore the importance of prioritising attributes that foster communication and collaboration in sustainable co-working space design for educational institutions. The study provides actionable recommendations for creating user-centric, sustainable co-working spaces that optimise human interaction.