Researchers have developed a new citation ranking system called EDM to identify and credit scientific breakthroughs. The system uses neural language models to provide stable representations of papers and can detect simultaneous discoveries, fixing a blind spot in the original disruption index.
Sociologist Mona Sloane argues AI's prediction paradigm reorganizes social relations and relationships. Her book 'Predicted: How AI Is Restructuring Social Life' advocates collective governance of AI as a form of infrastructure.
Researchers are developing next-generation adaptive insulin delivery systems using machine learning algorithms and ultra-rapid-acting U-500 insulin to personalize glucose control without manual intervention. The project aims to give people with Type 1 Diabetes their time, attention, and peace of mind back.
The University of Virginia has launched a cross-Grounds initiative to advance health and healthcare through data and artificial intelligence. The Collaboratory aims to accelerate discovery, share expertise, and lower collaboration barriers across disciplines.
A new clinical trial at UVA is testing an innovative AI-powered device designed to improve automated insulin delivery, aiming to simplify diabetes management. The trial will evaluate the 'Bolus Priming System with Reinforcement Learning' technology in adult participants with Type 1 Diabetes.
Peter Alonzi has been awarded a $78,594 grant to support the launch of the Criminal Justice Data Training Initiative at the University of Virginia. The program aims to provide undergraduate students with hands-on experience in data analysis and interpretation for local agencies.
The University of Virginia School of Data Science will receive a $20 million grant from the Quantitative Foundation to support the construction of a new building. The facility will enhance the school's research, scholarship, and teaching capabilities, while fostering an entrepreneurial ecosystem.
A team of UVA researchers, including Phil Bourne and Cam Mura, develop a new computational framework called DeepUrfold to explore structural similarities in proteins. This framework identifies faint relationships between proteins that were previously considered unrelated, revolutionizing the field of protein structure analysis.
A new study analyzed 3.6 million tax records to reveal the dynamics of philanthropic support for science in the US. The authors found that large donations are rare, and smaller donations drive most funding, with geography and shared priorities influencing giving patterns.
A UVA data science professor has received a $50,000 NSF grant to develop an automatic item generation and evaluation system for high-quality test items in K-12 standardized testing. The project aims to reduce laborious processes while ensuring fairness and alignment with required specifications.
Researchers analyzed 278 papers on motion capture technology, identifying six types of errors that reveal biases in measurement and validation processes across different eras. These findings highlight the need for critical examination of technological systems to prevent harms, such as car safety features affecting female injury rates.
Researchers found that fluvoxamine did not reduce symptom duration in patients with mild to moderate COVID-19, contrary to previous uncertainty. The study analyzed 1,175 participants and used a randomized clinical trial design.
Researchers propose network-based methodology to contextualize scientific publication impact, allowing for more accurate assessment of interdisciplinary discoveries. The new approach normalizes citation counts, mitigating biases in existing methods.