Articles tagged with Data Sets
A decade-long review of Kids First DRC shows how shared data expands research opportunities across diseases, institutions, and scientific disciplines. The resource has supported findings with potential to improve diagnosis, risk assessment, and treatment in pediatric care.
Jiaqi Ma's $660,307 grant aims to develop tools for understanding how individual components of training data affect large AI systems. This project will improve the performance and reliability of widely used technologies like language models and recommendation systems.
Researchers have developed an AI tool that can detect online propaganda in Kinyarwanda, a Bantu language spoken by 350 million Africans. The dataset, called KinyaProp, provides examples of misinformation in Kinyarwanda for large language models to learn from and recognize.
The VERITAS project establishes AI Assurance as a core function of scientific research infrastructure to document, review, and stress-test AI systems. It aims to develop practical methods for detecting vulnerabilities before they compromise scientific results.
FAIR² Data Management, powered by Senscience, has been named a 2026 CODiE Award Winner in the category Best AI Data and Analytics Solution. The platform enables researchers to turn data into trusted resources that can be discovered, built upon, and used responsibly with AI.
Researchers developed a photospike-based TRNG that harnesses unpredictable light-induced electrical charges to generate true random numbers. The device passed all 15 randomness tests and remained stable over millions of cycles, making it suitable for image authentication and deepfake detection.
Rates of new obesity cases rose by 4% between 2019 and 2025, with the largest increases seen among young adults. The study also found significant disparities in obesity rates across sociodemographic groups and geographical regions, highlighting the need for deep-seated change to address the underlying factors driving obesity.
A recent study found that the FDA approval of semaglutide for weight management led to a significant increase in calls to poison control centers, with over 8,000 cases reported by 2023. The majority of errors were preventable and stemmed from unintentional dosing or therapeutic mistakes.
A recent study analyzing 5.8 billion medicines dispensed in England from 2019 to 2024 found significant health disparities, with nearly twice as many medicines prescribed to the most deprived groups as those in the least deprived groups by age 40.
TurboLynx, developed by POSTECH researchers, analyzes complex, interconnected data up to 184 times faster than existing systems. The engine groups similar data together and processes them collectively, reducing unnecessary memory usage and enabling efficient analytical queries.
A recent study has identified five distinct species of pit vipers in the Himalayan region, including three new to science. The research combined fresh and historical DNA data with morphological and skeletal analysis, revealing deep evolutionary lineages and distinct ecological characteristics.
A recent study analyzing nearly 800,000 fashion images reveals that while representation has broadened, the typical female model body has remained remarkably stable. Diversity is primarily concentrated in a small number of models at the extremes rather than through a shift in the norm itself.
A new report from Brookings Institution highlights the federal government's growing use of AI, but also notes significant disparities and bottlenecks to widespread adoption. Large agencies lead the way, while smaller agencies struggle with workforce capacity and trust issues.
Researchers at King's College London have developed a way to overcome AI 'Data Cannibalism', a threat where AI models trained on generated data produce inaccurate results. By introducing a single datapoint from outside the closed loop, they can prevent model collapse and generate accurate results.
Researchers found that state-coordinated media in AI training data influences model responses about politics, especially in a country's own language. The team tested commercial models and found that adding scripted news to the training data made them produce more favorable answers.
Researchers from MIT have developed a more user-friendly and efficient method to identify potential system failures in cloud computing algorithms. The 'MetaEase' technique analyzes an algorithm's source code directly to uncover hidden blind spots that might cause unexpected failures, reducing the risk of costly network outages.
The UK-led OpenBind initiative has released its first publicly available dataset and predictive AI model, accelerating the discovery of new medicines using artificial intelligence. The release showcases high-quality, standardized experimental data and a trained predictive model, enabling researchers worldwide to drive the next generati...
Researchers found that camels have a more flexible and coordinated response to heat stress, allowing them to maintain stability even at higher temperatures. In contrast, human cells tend to respond in a more rigid way, making them less adaptable under heat stress.
Researchers at MIT developed a technique to overcome memory constraints and communication bottlenecks in federated learning, enabling faster and more accurate AI model training. The new framework, FTTE, uses a subset of model parameters and an asynchronous approach to reduce lag time and improve training performance.
MIT researchers have created an 'EnergAIzer' method that generates reliable results in seconds, allowing data center operators to optimize resource allocation and reduce energy waste. The tool leverages patterns from AI workloads and software optimizations to provide fast but accurate power estimates.
A new study developed China's high-precision, 1 km resolution soil moisture dataset using machine learning techniques. This dataset enables daily monitoring of soil dryness and wetness conditions across the country, providing critical support for drought early warning, flood forecasting, and agricultural management.
Researchers will use DNA-encoded chemical libraries and artificial intelligence to screen hundreds of millions of potential drug compounds, identifying those most likely to succeed in treating Alzheimer's. The project aims to shorten the timeline for identifying new treatments, bringing them to patients faster and with greater precision.
The Keck School of Medicine of USC and Tempus are creating a system-wide framework to integrate clinical care, clinical trials, and research through AI-powered precision medicine tools. The goal is to enhance patient care and accelerate research and innovation.
Researchers introduced QCell, a curated collection of 525,000 new quantum-mechanical calculations for biomolecular fragments. The dataset addresses the limited coverage of nucleic acids, lipids, and carbohydrates, enabling reliable simulations of critical biological processes such as DNA dynamics and membrane behavior.
A new Northwestern University study confirms that US drug overdose deaths have continued to decline following a peak in August 2023, contrary to speculation of manipulated CDC data. The study highlights the importance of accurate data for public health response and calls for greater transparency in federal data systems.
Researchers developed Sandook, a software-based system that tackles three major sources of performance-hampering variability simultaneously. The two-tier architecture optimizes task distribution for the overall pool while faster schedulers on each SSD react to urgent events.
The BrainHealth Network connects researchers across the country to understand brain health improvement through advanced MRI imaging and data analysis. The network leverages a comprehensive multimodal brain imaging dataset, including a longitudinal study of 100,000 healthy participants over 10 years.
Researchers examined how practice-based research and learning networks approach data governance, identifying the importance of building knowledge of Indigenous data sovereignty. Existing Indigenous governance frameworks provide guidance on incorporating Indigenous data sovereignty principles into PBRLNs.
The Shorebird Science and Conservation Collective uses big data to inform conservation efforts by analyzing tracking data from over 3,400 individual birds. The collective brings together data from various organizations to provide actionable information for land managers and decision-makers.
A global analysis of over 2,300 seawater samples reveals human-made chemicals make up a significant portion of organic matter in coastal oceans. Industrial chemicals, including plastics and consumer products, dominate the anthropogenic chemical signal, persisting even 20 kilometers offshore.
Engineers at the University of Pennsylvania have developed LIBRIS, an automated microfluidic platform capable of generating lipid nanoparticle formulations at high speed and scale. This enables the creation of large, systematic datasets needed to train predictive AI models, accelerating the design of lipid nanoparticles for mRNA delivery.
Dr. Dennis Lal has been appointed as the new executive director of the Center for Innovation in Health Informatics at UT Arlington, succeeding Marion Ball. He will lead initiatives on precision health, clinical AI, and health care-scale informatics.
Researchers from ISTA developed an algorithm that can extract and analyze information from the world’s most extensive biobank with unprecedented accuracy and speed. The method, dubbed gVAMP, enhances the framework's ability to extract complex information from the dataset at hand, providing a detailed overview of the effects on a trait ...
Researchers at UCSF and Wayne State University found that generative AI tools can perform orders of magnitude faster than human teams in analyzing health data. Junior researchers paired with AI generated viable prediction models in minutes, outperforming experienced programmers in hours or days.
A collaborative effort by researchers from the University of Göttingen and other institutions is creating a genomic inventory of European marine annelids. The goal is to accelerate biodiversity research worldwide and counteract the 'silent extinction' of marine species.
A recent NSF grant will support the development of new diagnostics and predictive models for understanding self-competition and weak asymmetry in turbulent flows. The project aims to uncover hidden patterns that current models miss, leading to improved simulations in weather forecasting, climate modeling, and engineering design.
The State of Open Data report shows open data has become embedded into research practice with FAIR awareness widely recognized. Researchers need systems that reward openness and workflows that make sharing effortless to sustain progress.
NEXMO Datahub is a mobility data space that enables secure and reliable data exchange between public and private organizations, fostering innovative solutions for smarter and more sustainable mobility. The platform aims to accelerate the digital transformation of the sector through data sharing among key stakeholders.
A study by Bielefeld University used anonymized WhatsApp metadata to show that personalized feedback can help people understand their communication habits. Many participants adjusted their views on response speed and chat participation after seeing data-based visualizations.
Researchers have launched a new multimodal medical dataset, Observer, capturing anonymized, real-time interactions between patients and clinicians. The dataset links video, audio, transcripts, and electronic health records to study subtleties like body language and environmental factors affecting care.
A new study maps three decades of income inequality data globally, revealing worsening trends for half the world's population but 'bright spots' in regions with effective policies. Regional efforts such as investments in public health and education in India and cash transfer programs in Brazil show promise in reducing inequality.
Researchers analyzed AI methods for detecting pulse rates from facial video recordings and found significant errors at elevated heart rates. The study highlights weaknesses in remote photoplethysmography (rPPG) technique under challenging conditions.
IMDEA Networks has created a secure watermarking tool called FreqyWM, allowing institutions to tag their data with a unique signature and preventing leaks and unauthorized copies. The tool enables the exchange and reuse of information while complying with legal frameworks on security and privacy.
A new report demonstrates how harnessing digital data from mobile phone applications and social media platforms can provide faster, more spatially fine-grained estimates of population movement. This information is crucial for delivering timely humanitarian assistance during crises.
The team created a specialized two-dimensional thin film dielectric designed to replace traditional heat-generating components in integrated circuit chips. This breakthrough aims to reduce the significant energy cost and heat produced by high-performance computing necessary for AI.
A new open-source tool, SCGraph, enables users to calculate the shortest connection between two points worldwide across different transport modes. The system is based on a novel shortest path model that integrates data from road, rail, and maritime routes.
A team of researchers from Binghamton University has created the first-ever high-resolution 3D model of Rano Raraku quarry, revealing over 1,000 moai statues. The model allows users to zoom in and pan across various features, providing a detailed look at the island's quarries and challenging previous theories about its history.
A comprehensive dataset detailing Japanese adults' responses to the pandemic offers unprecedented insights into public attitudes and behaviors. The 30-wave panel survey captures how risk perception, preventive behaviors, and psychological distress evolved over four years.
The Home Health Focus dataset provides insights into Medicare home health use from 2016 to 2019, including demographic data and patient function indicators. The analysis shows a rise in home health stays among beneficiaries while a decrease in active agencies during the same period.
The algorithm identifies the minimum set of locations where field studies would guarantee finding the least expensive route, considering problem structure and uncertainty. This method can be applied to broad classes of structured decision-making problems under uncertainty.
Pusan National University researchers develop a novel prompting technique to improve ChatGPT's accuracy in predicting fashion trends. The study reveals that ChatGPT can capture emerging themes and identify new trends not found in existing data.
The Stowers Institute has appointed its first AI Fellow, Sumner Magruder, to harness the potential of artificial intelligence in biological research. He will collaborate with researchers to design new algorithms and unlock insights from large datasets.
The Global Pathogen Analysis Platform (GPAP) will enable low- and middle-income countries to conduct research and surveillance of infectious diseases independently. The platform aims to prevent disease outbreaks from developing into pandemics by detecting genetic sequences of potential pathogens.
The African Data Drive is an interactive tool that provides accessible, quality-assured spatial data to empower decision-makers to balance development needs with conservation priorities. The platform enables users to assess potential risks to biodiversity and access the most appropriate information on sustainable development.
A new project aims to enhance workforce readiness in molecular bioscience by creating open-access resources and modules tailored to student needs. The Molecular Data Education Hub will host instructional materials and case studies for instructors to implement into their courses.
The Variant Workbench enables researchers to explore genetic data in a single, integrated workspace, linking genomic information with clinical conditions. By reducing data complexity, the tool facilitates scientific discovery and accelerates pace of research.
Researchers at the University of Missouri have developed an AI-powered method to detect hidden hardware trojans in chip designs, offering a 97% accurate solution. The approach leverages large language models to scan for suspicious code and provides explanations for detected threats.
Frontiers' revolutionary AI-powered FAIR² Data Management service is transforming the way scientific data is shared, enabling researchers to fuel progress and earn credit for their work. The service ensures every dataset is preserved, validated, citable, and reusable.
The BIRDBASE dataset covers 78 ecological traits across 11,589 bird species, revealing that 54% are insectivores, with many tropical forest species under pressure. Fruit-eating birds disperse seeds in tropical forests, and fish-eating seabirds face elevated extinction risk.
A new framework for flexible data collection has been developed by Masumi Hisano, allowing for counts to take place in various settings, including cities and daily routines. This approach can help increase sample size and provide valuable insights into bird species assemblage datasets linked to landscape characteristics.