Repository Platforms for Research Data BoF

Hi everyone,
Thanks to all who attended the Repository Platforms for Research Data BoF at last week’s RDA Plenary in Amsterdam. The session was well-attended (over 100 participants) and we had a good discussion about starting a working group on this topic. I have included notes from the meeting below; thanks to Stefan Kramer and Sarah Jones for taking notes during the meeting.
I believe we have sufficient interest to begin the process of establishing a working group. The group will collect and analyze research data use cases in order to produce a set of functional requirements (tied to the use cases) for repository platforms. There was general agreement that the target audience for this output will be developers of repository platforms, along with service providers. Secondary targets could include funding agencies and anyone trying to map their institutional use cases to functional requirements.
I will send a follow-up email to those who left me their business card (or who contacted me directly about the group). If you have not given me your contact information but would like to join the group, please get in touch with me so I can add your name to the list. 
The next step is to draft a case statement [1] to formally create the working group. I will get started on this and post a draft to a wiki or similar shared space so anyone interested in the group can help work on it.
—
Notes
* Goal: Start a Repository Platforms for Research Data working group spawned from the Long Tail of Research Data interest group
* Collect research data use cases, analyze in the context of repository platforms, produce recommendations on functional requirements
* Who would these recommendations be intended for, what problem would be solved by this WG? How can we distill some of these problems into functional requirements?
* Target audience: existing or potential future developers of repository platforms e.g. DSpace, Fedora developers
* There may be many use cases that we, and repository developers, are not aware of, and that therefore warrant gathering.
* A WG could act as a conduit to researchers to gather needs, though (arguably) researchers may not be regarded as "the end users" of repositories, while libraries stand in between the researchers and repositories (also arguable). Could we have architectures that bridge the gap ... which we don't now?
* A WG should understand researcher needs if they are expected to be direct data depositors.
* Should scope be extended to include "smart grids" or specific research workflows?
* It would be helpful to know what repository platforms work well today.
* The functions and scopes of repositories can vary, and this WG should define some of those parameters.
* Functional requirements could focus on developers and generic APIs for interoperability.
* How does "big data" impact the repository paradigm?
* It's not just about final/archival storage, but also limited-lifespan backups.
* Service providers that offer repository services could be another target audience.
* Being effectively transparent to users should be an element of a multiple-repository/platform architecture development, so interoperability is an important aspect to look at.
* Limiting the use cases to work against would be important for the WG to remain practical.
* Funding agencies could perhaps be a potential secondary target audience. Example: DOE wants community input, not prescribe what repository platforms researchers should use.
* Commercial service providers might be another target audience, to integrate needs of the academic community.
* Repository operators might be another target audience, to understand how well their repositories meed researcher needs.
* The WG would first have to define a use case structure/format.
* Review of related work would be important. CKAN has created a spreadsheet of requirements. Communication/coordinating with other RDA WGs would need to be undertaken. Communication within WG would take place via RDA mailing list.
* Adoption plan would include how to reach the intended target audiences, with partial goal of getting functional requirements adopted.
* Structure of WG could include co-chairs, or chair and vice-chair.
* RDA PID working group has developed an API - as an example of actual WG output from RDA.
* The Open Repositories conference might be a good place for a WG meeting; however, few attendees in the room had attended the last one.
* Putting up a wiki for stakeholders to submit use cases could be a possibility; worked well in the Fedora community. Direct personal contact with individual stakeholders and taking a survey could be a good approach.
* The WG members should reach out to their own communities to gather use cases.
* Use case gathering should not just be a single open-text submission.
* Use case "ghost writers" may have to be engaged, because researchers won't write them themselves. This has been done successfully at one organization in the USA, which generated about 80 use cases this way. www.xsede.org - John Towns
* It may be more important to analyze what use cases are already out there than spend time with gathering new ones. This could be part of the "literature search" by the WG. However, efforts are already underway to gather use cases from scientists.
* How wide of a net are we going to cast in terms of use cases for a manageable 12-18 month project is something the WG will have to decide upon.
* WG participants need to understand it will require a time commitment to produce useful output.
* WG should define existing and required functional requirements of repositories, and also work on organization's needs.
* A matrix of use cases vs. funct. requirements could include funder/regulatory requirements. That could very important for repository developers.
* Invitation to join WG should also be spread beyond RDA.
—
Regards,
David
[1] https://www.rd-alliance.org/working-and-interest-groups/case-statements....
-- 
David Wilcox
Fedora Product Manager
DuraSpace
***@***.***
Skype Name: david.wilcox82
From: David Wilcox <***@***.***-commons.org>
Reply: David Wilcox <***@***.***-commons.org>>
Date: September 21, 2014 at 8:04:53 AM
To: ***@***.***-groups.org <***@***.***-groups.org>>
Subject:  Repository Platforms for Research Data BoF
Hi everyone,
This is just a reminder that the Repository Platforms for Research Data BoF will take place on Monday at 11:00am at the RDA Plenary. You can find a full description, along with an agenda for the meeting, on the RDA website [1]. Anyone interested in exploring this topic and potentially joining the forthcoming working group is welcome to attend. Ideally we will lay out some goals and objectives for the working group, make a plan for creating the case statement, and discuss the process of gathering and analyzing use cases. Please let me know if you have any questions.
I’m looking forward to the meeting!
Regards,
David
[1] https://www.rd-alliance.org/repository-platforms-research-data.html
-- 
David Wilcox
Fedora Product Manager
DuraSpace
***@***.***
Skype Name: david.wilcox82