Attendees

• Jonathan Markow, DuraSpace
• Carissa Smith, DuraSpace
• Susan Parham, Georgia Tech
• Chris Helms, Georgia Tech
• Joan Parker, Moss Landing Marine Labratories
• Sid Byrd, Rice University
• Ying Jin, Rice University
• Gail Steinhart, Cornell University
• Deb Morley, MIT
• Brian Westra, University of Oregon
• Mike Wright, NCAR

Introductions

• Some folks attended requirements workshop in October, some are representatives from institutions that attended, and some are new today:
  • Gail Steinhart, Cornell University
  • Deb Morley, MIT Libraries (sitting in for Steve Gass), looking at various research data service opportunities
  • Susan Parham, Georgia Tech and Chris Helms network administrator
  • Sid Byrd and Ying Jin, Rice University (sitting in for Geneva Henry)
  • Mike Wright (NCAR)
  • Joan Parker, library/it/data management at Moss Landing Marine Labratories
  • Brian Westra, University of Oregon, science and data librarian

Discussion

• Summary of activity since last meeting
  • general feedback from participants was that it was very helpful
  • took discussions from meeting and developed use case stories for various roles
  • then created a list of what the system should do based on those roles and then started to prioritize (and also scope of first phase)
  • first phase of DfR ends at end of calendar year 2012
  • also created a narrative of the scope of items/activities the system will perform this year
  • then developed high level architecture to support the system
  • established partnership with Smithsonian and will be creating a data management and visualization system designed to look at research data that is stored in a Fedora repository and create incentive for researchers to add metadata to data for them to manage, organize, share data
  • DfR is partnering with Smithsonian to include their work for visualizing and managing data
  • have also identified four iterations of agile development for the year that also prioritizes tasks
• Summary of narrative report
  • basing DfR system on DuraCloud (which provides preservation services in the cloud that integrates with public cloud providers Amazon and Rackspace, as well as other providers in the near future)
  • DfR extensions will preserve researcher data which is very vulnerable from the beginning (not institutionally supported, ad hoc)
  • DfR will grab data from very beginning (regardless of where it resides), ask researcher what should get backed up, and give them options to add metadata about the data
  • question from Deb: how does this play out when there are multiple researchers, as well as researchers from multiple organizations
  • answer: DfR will support groups and have a management console where many people can have access to data
  • question from Deb: does local copy of data need to be maintained locally?
  • answer: we must be able to accommodate both scenarios, our initial expectation is that it would be a backup
  • a Fedora repository will be the backend of the Smithsonian application for managing and visualizing data
  • data in DuraCloud, metadata in Fedora (that is backed up to DuraCloud), web interface via Smithsonian application
  • Smithsonian application allows you to look at your data, describe the data, describe the relationships between the data, define your project concept, define associated entities and sub-projects, all of this will allow for the automatic creation of a project web site
  • question: transitory system not for longterm preservation?
  • answer: not sure what choices will be made from institutional level, but data is being stored for longterm, but migration options will also exist
  • question: are there any demos of the Smithsonian application available now or any upcoming?
  • answer: no public demos, just presentations; application is just getting to the stage where it can be demoed; they are trying to make that happen at Open Repositories conference and at some point we will install a version of the interface into the DfR system and we will be in the position to demonstrate it
  • question: future development on repository system and what tools would be enabled
  • answer: development team working on Smithsonian application has a lot of experience creating researcher web applications, so we hope that the application will evolve into a researcher environment in the cloud
  • Internet2/InCommon shibboleth integration that will enable authentication and authorization
  • feedback: intrigued by web site output service and assuming it will tied into datacite for DOIs
  • answer: we will integrate with a number of third party services that support researchers (datacite will be one); also identifiers, data management planning tool (from CDL), create a template plan
• Feedback from researchers is essential
  • researcher input in regards to adding metadata, interacting with the interface, the user interface to the system is greatly needed
  • email Jonathan if you would like to participate
  • feedback/recommendations for moving forward
  • feedback from Gail: take show on the road to scientific conferences to get feedback
  • recommendations welcome to those types of conferences, emails welcome
  • feedback from Joan: what would you do with live datastreams such as those coming from instrumentation directly
  • not something we will be able to achieve in the first iteration, because of difference between devices and proprietary software associated with those devices
  • question regarding preference of disciplines or metadata structures?
  • answer: we would like a variety of disciplines, but will have to limit quantity of data for testing, we will set parameters to data sizes, number of files, etc.; first round of testing will focus on usability of user interface and application
  • question: role of curator toward the end of the process, envisioning any other roles of library staff to engage with content from a curatorial perspective
  • answer: will vary based on institution, there will be a continuum of partnerships around this
  • suggestion: datacurationprofiles.org would be helpful as a resource, they provide a toolkit with questions about metadata and how to characterize data
  • send email if you are interested in having researchers and/or other ways to volunteer