Post on 04-Jan-2016
How many sizes are needed to fit all research data?
Todd Vision
Associate Director for InformaticsNational Evolutionary Synthesis Center
Associate Professor of BiologyUniversity of North Carolina at Chapel Hill
ESIP Summer 2013, Chapel Hill, NCData Decadal Survey Panel
© AAAS
My impeccable credentials
• Not an Earth Scientist• Blissfully ignorant of whole
Decadal Survey process until a few weeks ago!
My nonetheless strongly held opinion
• We need a national roadmap for investment in basic science data infrastructure
• The decadal survey process seems as good a way to go about it as any
The OSTP public access memo is a rare
opportunity• The desiderata for data are well
articulated• But funding must be found from
within existing budgets • And mechanisms to ensure
interagency coordination are weak• Soon we will have approximately
20 responses to evaluate and compare
The issues cut across disciplines
• Balancing infrastructure, research investment, workforce development
• Leveraging commercial innovations
• Coordinating across agencies• Coordinating w/ international
efforts• Ensuring buy-in from the research
community• Assessing success
Differences within disciplines are
mirrored between them
• Data assets range from reference collections to one-offs
• Things vary predictably along that continuum Approaches to data curation Preservability Return on investment in specialized
infrastructure
Narrow scope = narrow impact =
narrow opportunities• Unique feature of the moment is
that all research agencies will be laying their cards on the table
• Unique opportunity to identify cross-cutting investments,
priorities receive support from multiple
agencies to build on OSTP and RDA’s
momentum
Target actions that agencies can take
• Investments and, secondarily, policies
• Can not no so easily guide the actions of researchers, research institutions, societies, journals, etc.
• Individually, research community voices tend toward private benefits
• A common voice could have tremendous influence on government actions
Research is international
• A survey of US policies and investments in relation to other countries would be enlightening
• What opportunities are being missed?
• What mechanisms exist for greater international cooperation?
• Where can and should the US ‘compete’?
Governance and sustainability
• Are there ways to ensure future interagency cooperation?
• International cooperation?• How to combine infrastructure
funding with private investment and continued funding for infrastructure innovation?
• How to ensure responsiveness to changing practice?
Risks
• Not adding value beyond disciplinary efforts like EarthCube
• The temptation to propose some politically attractive megaproject
• Forcing premature consensus where experimentation is needed
• Suppressing grassroots initiatives
What does success look like?
• The process itself needs to consider what ROI we should maximize
• Is there Buy-in to the process and
conclusions? Investment that follows
recommendations? An increase in cross-agency, cross-
disciplinary collabocoordination?
What is missing from this
conversation?• Discussion of data in the context
of Software Physical collections Scholarly literature
We need a national roadmap for investment in basic science data infrastructure.
The decadal survey process seems as good a way to go about it as any
Image credits
• Data wordle: © Yael Fitzpatrick and AAAS, http://www.sciencemag.org/site/special/data/ScienceData-hi.pdf
• Butterfly Alphabet © Copyright 1965-2011, Kjell B. Sandved
• Highway interchange: CC BY-SA 2.0 avlxyz, source:http://www.flickr.com/photos/avlxyz/4589977933/