Making all research data accessible, discoverable, and usable
Globus Labs is a research group led by Prof. Ian Foster and Dr. Kyle Chard that spans the Computation Institute, Department of Computer Science, and Math and Computer Science Division at the University of Chicago and Argonne National Laboratory. Our modest goal is to realize a world in which all research data are reliably, rapidly, and securely accessible, discoverable, and usable. To this end, we work on a broad range of research problems in data-intensive computing and research data management. Our work is made possible by much-appreciated support from the National Science Foundation, National Institutes of Health, Department of Energy, National Institute of Standards and Technology, and other sources, and in addition to computer science, engages fields as diverse as materials science, biology, archaeology, climate policy, and social sciences. We work closely with the team developing the Globus research data management platform, who often challenge us to think bigger—and sometimes implement our less crazy ideas.
We are developing methods to automatically extract scientific facts buried in scientific publications
Logan Ward gave a talk on Materials Database at the Workshop on “Gap Analysis: Materials Discovery through Data Science at Advanced User Light Sources” in Santa Fe, New Mexico.
Yuliana Zamora presented a poster at CERES titled “Data Driven Analysis of GPU Benchmarking”. The poster describes a data-driven framework that makes use of data-analysis tools, such as machine learning and transfer learning, to reduce the time needed for engineers and researchers to predict the performance of a given software kernel on an arbitrary GPU architecture.
Roselyne Tchoua presented a poster at CERES titled “Towards Hybrid Human-Machine Scientific Information Extraction”. The poster describes different combinations of human-machine approaches to extracting scientific facts from the literature, specifically polymer names and properties from a relevant scientific journal in polymer sciences.
Ian attended the 2018 Workshop on Clusters, Clouds, and Data for Scientific Computing (CCDSC) in Lyon, France. He presented on the computer systems challenges raised by the rise of deep learning in science, and technologies that Globus Labs is developing to address those challenges.