Quantifying FAIR: metadata improvement and guidance in the DataONE repository network
Speakers
Peter is a Software Engineer at the National Center for Ecological Analysis and Synthesis (NCEAS) at the University of California, Santa Barbara. He has contributed to projects at NCEAS including DataONE infrastructure, data processing provenance (capture, archival and display), metadata quality, and scalable computing systems.
Matt directs the Informatics program at NCEAS, which focuses on both supporting efficient synthesis through scientific computing and on building new advanced infrastructure to support data sharing, preservation, analysis, and modeling. Matt is the Director of the DataONE program, a global network of interoperable data repositories, and of the NSF Arctic Data Center. In addition to data infrastructure work at NCEAS, Matt also helps to build the NCEAS Learning Hub through an emphasis on data science and reproducible research teaching.
Matt’s career has focused on improving data science infrastructure to support cross-disciplinary and synthetic science, principally through the development of open source software for data repositories, metadata systems, and reproducible analysis and modeling.
Matt has a M.S. in Zoology from the University of Florida that focused on the ecology of plant-animal interactions, and a B.A. from Dartmouth College.
DataONE has consistently focused on interoperability among data repositories to enable seamless access to well-described data on the Earth and the environment. Our existing services promote data discovery and access through harmonization of the diverse metadata specifications used across communities, and through our integrated data search portal and services. In terms of the FAIR principles, we have done a good job at Findable and Accessible, while as a community we have placed less emphasis on Interoperable and Reusable. We will present new DataONE services for quantitatively assessing metadata completeness and effectiveness relative to the FAIR principles. The services produce guidance for FAIRness at both the level of an individual data set and trends through time for repository, user, and funder data collections.