DDI Active Data Management Plans Working Group Meeting - 13 December 2016

Attendees: Irena Vipavc Brvar, Churck Humphrey, Leahey, Jared Lyle, Sarah Jones, Anja Perry, Christina Guldfeldt, Maude Jones and Steve McEachern

The call focused on next steps for the group. Decided on two levels of activities:

1) Flow of information from DMP tools to archives. Components include:
a. APIs supported by tool providers to allow us to capture information
b. Mappings to use information in DDI specification
c. Assigning DOIs to DMPs that are published. How to incorporate persistent IDs into metadata captured.

Next steps: Selected archives can work with DMP Online/Tool on pilot project. It was agreed that Jared will organize a few of the social science data archives to implement an exchange of DMP content from DMPOnline with an automated mapping of themes to DDI-Lifecycle elements (Sebastian's work). They will look at the upcoming API implementation for DMPOnline to see how themes might be packaged for exchange (e.g., using JSON-LD to transport DMP text) and then processed into DDI. They will also consider how PIDs might be integrated into this process, including the possibility of DOIs.

2) Better understanding of what we mean by active -- multiple phases across project, or machine actionable activities? Requires standards developers and tools supporters to have discussions.
a. Active DMP has metadata-driven activities (e.g., storage question, which pushes out storage report that gets aggregated into institutional requirement for storage).
b. Look at how model team can support active DMPs. Develop use case for DDI developers to look at. Then once we have that we can get program developers together.

Next steps: 1) Take advantage of IDCC workshop to identify appropriate use cases. These use cases should address the meaning of "active DMPs". Two examples were given: (i) active in the sense of maintaining a record of RDM activity across the project (e.g., the use of Phases in the current implementation of DMPOnline); (ii) active in the sense of metadata driven RDM activities (e.g., notifying IT the need for storage). Check if we can get 1-2 model specification DDI developers to IDCC workshop. 2) Try to be more of a leader in setting the direction of RDA ADMP interest group. Use cases developed at IDCC workshop can provide concrete working activity for the RDA working group. Internal to DDI, we can propose ADMP be tackled during the May 2017 sprint.

Follow-up call to be scheduled for March 2017.

DDI Active Data Management Plans Working Group Meeting - 14 November 2016

Attendees: Irena Vipavc Brvar, Amber Leahey, Jared Lyle, Stephanie Simms, Jessica Gallinger. Maude Jones and Chuck Humphrey attended an earlier call. Via email: Sarah Jones

The primary focus on the call was learning about activities of the DMP online tools and how DDI could collaborate, especially so DMPs could interoperate and be reused.

The DMP online tools classify important items in data management plans using "themes", which represent the most common topics addressed in DMPs. Recently, Both Sarah (https://dmponline.dcc.ac.uk/) and Stephanie (https://dmp.cdlib.org/) have mapped all the funder requirements, tagged these with DMP "themes", and condensed these down to core points (n=17) with some of the original themes combined into broader categories (e.g., Data Selection + Preservation Plan + Period of Preservation = PRESERVATION). They have a proposed list out for consultation: http://www.dcc.ac.uk/blog/common-set-themes-dmps-seeking-input

From Stephanie: "The idea is that they should provide a basic tagging structure to facilitate text mining of DMPs, in addition to making it easier to attach customized guidance across templates. A larger goal is to coordinate this basic vocabulary with a more specific vocabulary(-ies)—CASRAI has already drawn from the themes for the DMP Ethics vocabulary (see http://dictionary.casrai.org/DMP_Ethics_Review) and we hope they will also prove useful for the DDI working group. To this end, we're delighted about the timing of our activities and look forward to incorporating any and all community feedback. With regard to our co-development efforts, we're focused on migrating DMPonline to the new platform as soon as possible since we're starting from their original codebase. The DMPTool migration will come later, after we've added all of the features from the MVP roadmap <https://github.com/DMPRoadmap/roadmap/wiki/Development-roadmap>. We estimate that DMPonline will be using the new system by Q1 2017 and DMPTool should be shortly behind in Q2 2017. Weiwei Shi—who manages development on DMP Assistant, as you may already know—has forked the new codebase so she'll be in the loop regarding new development and releases. We can implement the revised list of themes sooner or later, certainly by the DMPTool migration, depending on how long it takes to gain buy-in from this and other DMP-focused working groups (RDA, Force11, etc)."

The group discussed next steps. Decided it would be good to review the 17 DMP core "themes", determine if these make sense and if other items should be included, and then map these items to DDI 2 and DDI 3. There are still gaps in what DDI captures before data are gathered (see: http://blogs.ucl.ac.uk/dmp-ss/files/2013/05/Data-Management-Planning-and-the-Data-Documentation-Initiative_1_0.pdf) and we should document those gaps and propose additions, where needed. Once the review is completed, we can discuss further with Stephanie and Sarah about next steps, including establishing a metadata model for DMPs.

TO DO:

1) Several members of the group are planning to meet in person at the European DDI (EDDI) user conference in early December. This would be a good opportunity to review the DMP core "themes" and complete mappings. If you'll be at EDDI and would like to meet, please complete this Doodle poll by this Wednesday (16 November):

http://doodle.com/poll/reyu7dvx3q2wdash

2) We'll also plan a follow-up call for mid-December (I've tried to be sensitive to worldwide time differences – please indicate all times that would be realistic for you and we'll select the time that works for the most people). If you'd like to join the call, please indicate your availability via this Doodle poll by this Friday (18 November):

http://doodle.com/poll/668c2d56kadie8iq

Note that Stephanie emailed after the call with the following additional information:

Below is a list of resources that you're welcome to circulate and/or post to the DDI wiki. We're happy to address questions about anything herein via email or on future calls.

I also want to invite someone from this group to participate in a workshop we're hosting at IDCC (International Digital Curation Conference) on 20 Feb (details attached). Developers are our target audience and capacity is limited owing to room sizes, but we'd like to have a representative group of DMP-type people and places. Please let us know if anyone plans to attend. 

Resource links:

• Blog post about DMPonline themes http://www.dcc.ac.uk/blog/common-set-themes-dmps-seeking-input
• Original set of 28 themes and mapping to UK funder questions http://www.dcc.ac.uk/sites/default/files/documents/publications/DMP-themes.pdf
• Revised set of 17 themes http://www.dcc.ac.uk/sites/default/files/documents/publications/DMP-themes-revised-Sept16.pdf
• Machine-actionable DMPs slides http://www.slideshare.net/StephanieSimms/making-dmps-actionable-and-public
• PIDapalooza slides for "PIDs in DMPs" (pardon the music puns) http://www.slideshare.net/StephanieSimms/pids-in-dmps-spinning-tracks-with-syntax
• Utah State Univ DMP in Digital Commons repo http://digitalcommons.usu.edu/funded_research_data/
• Utah State Univ paper and slides about DMP pilot project http://digitalcommons.usu.edu/lib_present/90/
• DMPRoadmap (merged platform) GitHub repo with code and some project info on the wiki https://github.com/DMPRoadmap/roadmap

Currently, we're focused on developing the new, merged platform code-named DMPRoadmap (link above) — this will occupy all DMPTool/DMPonline resources until spring 2016. The themes will be part of the new platform for both tools, but other functionality related to machine-actionable DMPs will come as future enhancements. We're also actively seeking funding to move things forward more quickly. Regular (monthly) status updates are available on both DCC and DMPTool blogs.

https://blog.dmptool.org/

http://www.dcc.ac.uk/drupal/blog

Again, please let us know if you have any questions and/or feedback.

DDI Active Data Management Plans Working Group Meeting - December 15, 2015

Attendees: Chuck Humphrey, Amber Leahey, Irena Vipavc Brvar, Iris Alfredsson, Maude Frances, Tito Castillo, Uwe Jenson, Michelle Edwards

Regrets: Anne Sofie Fink,

• Introduction by the Chair and a recap of the Terms of Reference and mission of this working group
• Introductions of group members
• There are different perspectives of a DMP - driven by demands of different countries and different funding agencies and requirements
  • European vs other countries
  • within Europe - differences exist
• RDA - DDI proposal
  • Mary attended last year - at that moment there appeared to be no coherent approach to DMPs
• So many different approaches to DMPs
• Some have mandatory fields
• some are optional
• ideally we have to ensure that the context is the same
• What is missing from DDI now
• want to capture in advance
• how to package and exchange with other administrative information/units
• some elements should be mandatory - elements that are shared with other systems
• be able to compare within and across systems
• How does DDI inter-operate with other standards
• DDI - CRIS?
• HR & Grants management
• Essex system and Fedora - will be linked to data discovery
• early stages
• needs to be a common service across different documentation management systems
• no discovery layer for DMPs at the moment
• In Germany, DMPs are very vague at the moment - no set criteria
• Might be a good idea to do a survey of DMP structures among countries
• This would feed the DDI fields that are needed
• DDI value = provides structure and comparability
• could answer questions such as:
• how has a plan changes (Comparison module could play a role here)
• added value - computable DDI
• audit trail vs. accountability
• Documentation of Use Cases is crucial here!
• Data security

DDI Active Data Management Plans Working Group Meeting
September 10, 2105

Present: Chuck Humphrey, Amber Leahey, Steve McEachern, Leanne Trimble, Mary Vardigan

Invited: Amber, Arofan, Chuck, Leanne, Mary, Michelle, Steve, Stuart, Tito

1. Welcome new members

1. Steve, Stuart, Tito

2. Minutes of the 28 August 2015 meeting

1. Business arising from the minutes

1. Status of the recruitment of members to the ADMP DDI working group

Should we contact the DMP Online developer groups? We have some overlap with DMP Online through the Canadian contingent, Portage, and the DCC. The DMP Online is developing an API and we should try to learn more about that. There is no existing metadata model behind the DMP Online so they are looking at querying the content using the API. They understand that having a metadata model would help exchange of information and interoperability. If we make our interests known to them, we might be able to work with them. Steve will also help us reach out to ANDS about involvement in the group. Action: Reach out to the two organizations.

ii. RDA activities

We have not yet heard from the ADMP group about getting a slot in the session, so we are awaiting a response. There have been some negative responses from researchers about the fact that the current DMPs are not active but static and not reusable. There is a burden on the researcher, but many other stakeholders can use this information.

1. White paper and Dagstuhl

Amber has started on a white paper about approaches in DDI to develop metadata support for Active Data Management Plans. The paper still needs work. Action: Chuck and Amber will talk offline about the goals of the paper and how to coordinate the writing. This is a great use case for interoperability across metadata standards -- especially between CASRAI and DDI. How can we bring standards together in an interoperable way? This will be of interest to people at Dagstuhl attending the Metadata Workshop.

1. A discussion about three approaches to ADMP

1. The three identified in Arofan and Tito’s paper

1. Incorporate DMPs into the DDI schema -- create new fields, tags, relating to DMPs

2. Reference a DMP schema -- add a reference or link to a DMP document or schema  

3. Interoperate with another schema that describes DMPs -- map to another schema (superschema) -- that is, think about DMP information outside of DDI and identify information inside DDI that would support DMPs

Arofan and Tito’s paper says that the current DDI standard isn’t sufficient to describe DMPs. If we went with i, we could incorporate it as a DDI profile. Codebook probably wouldn’t be appropriate as it is describing information after the fact. Lifecycle is most appropriate. How would DMPs fit into the DDI model approach? Yes, we could then work backward into Lifecycle. Having a model helps other communities and domains understand what we are doing.

The paper recommended that there be conceptual mappings between DMPs and the current Lifecyle. The Methodology section of DDI is messy and is out of date and needs refreshing. That is the one caveat about starting from existing fields.

1. Mechanisms or tools

1. Beacons or buttons

        We don’t want the technology driving the content, but it is hard to envision how the content will be expressed without tools. How might an application make use of it? DMPs are structured around templates and questions and that makes it difficult to have a governing underlying structure. It’s administrative information. One strategy is that it is built around the exchange of information so it’s an exchange protocol. Bringing Arofan in about the modeling approach would be helpful. This would allow our standard to be easily used by different information technologies.

Next meeting: Early October

ADMP Working Group Meeting
August 28, 2015

Present: Michelle Edwards, Amber Leahey, Leanne Trimble, Mary Vardigan

Approval of ToR

The group suggested adding an item about communicating and sharing our work with other relevant groups. This was added as a seventh goal: Communicate about this work to relevant organizations (e.g., RDA).

Recruiting Members

Two names had been suggested from initial contacts in Germany, and the group also suggested people from the UK and Australia. Mary will follow up with invitations.

Outreach to RDA

There will be a joint meeting of three relevant interest groups at RDA: IG Preservation e-Infrastructure, IG Active Data Management Plans & IG RDA/WDS Certification of Digital Repositories. It was suggested that we try to get 10 minutes on the program to discuss our ideas about a metadata model for ADMPs based on DDI. Mary will be in attendance and will request a spot on the program.

White Paper and Next Steps

The group was enthusiastic about Amber’s start on the white paper and will plan to carry this forward with the goal of presenting it to the external experts at Dagstuhl in the third week of October.

There were three ideas put forward in the original paper by Arofan and Tito: Incorporate DMPs into the DDI schema; reference a DMP schema; or interoperate with another schema that describes DMPs. We can explore these ideas further in our paper.

Present: Michelle Edwards, Chuck Humphrey, Amber Leahey, Mary Vardigan

1. Welcome to this Working Group

The group hopes to expand in size in order to accomplish all of its objectives. We hope to achieve our outcomes in a year to 18 months.

1. Define the Terms of Reference and Goals

1. Outline for ToR

1. Purpose of the WG

1. Identify the role the DDI Alliance will play in international Active Data Management Plans (ADMPs)

1. Need for standards with DMPs

1. Establish the DDI Alliance as a collaborator (player) in the ADMP space

1. Interoperability with other standards and the tools based on them

2. Mechanisms to draw in other standards

1. Provide guidance to DDI specification development incorporating relevant ADMP elements

The discussion at the Members Meeting that prompted formation of this group did not identify a purpose per se. We recognize the need for standards in data management planning. There has been some research into this early on with Tito Castillo in the UK.

Tito’s group didn’t see a role for DDI in this, but this working group thinks DDI is very appropriate for DMPs. There is an important metadata aspect to this and DMPs are part of the lifecycle. The information in DMPs evolves and can be reused and repurposed. The “active” aspect of this is also very important and changes the nature of the conversation.

We need to identify a role we can play in this space in terms of standards. We should be looking at reframing the earlier research to show DDI’s relevance.

The Alliance can be a collaborator in this space. The bridging of administration information and the actual research is important. The standards should be interoperable. Crosswalks are key to showing commonalities.

The DMP Tools do not use standards as the back end. Generally software developers have not used underlying metadata models, but DDI has turned this on its head. Different DMP tools should be able to share standardized information. The information in DMPs should also be in the metadata record for preservation. This should be machine-harvestable.

We could invite someone from DART at Oregon to join us; they have done research on DMPs. CISER is working in this area as well to determine data availability.

The WG agrees on positioning the Alliance as a collaborator in this space. Providing guidance internally to our own specification development as well as guidance to the wider community are important goals.

1. Goals and Deliverables

1. Reframe earlier research by Tito saying DDI wasn’t appropriateto show DDI’s relevance to DMPs

1. Contribute a mapping between DDI specifications to ADMP elements [white paper]

2. Recommend ADMP elements for consideration in DDI-M [statement]

3. Recommend exchange methods between ADMP and DDI online services [statement]

Can we contribute a mapping between DDI and DMPs? Our description of files in DDI-C maps to the files section of a data management plan. The identification of metadata is asked about in the DMP and this relates to Section 1 of DDI-C. Such a mapping would support the reframing of the research. We could showcase something tangible and write a white paper on this to address DDI’s role in this area. Ethics and licensing are other overlaps. Mary has a paper that she and Libbie wrote on this topic that she will share with the group as a possible starting point.

Another deliverable might be exchange strategies for interoperability. We may need to push this discussion along by serving as a use case for that kind of development work. The Alliance looked at the XACML -- Extensible Access Conditions Modeling Language -- in this regard, and the issue that came up was whether the XACML was being actively maintained and updated.

JSON and JSON-LD are becoming a common exchange format and APIs are using that to package information being exchanged through Web services. APIs using only JSON are limited, though, and there may be a need to explore this. We should promote the use of existing standards like OAI for harvesting information.

1. Membership

1. Regional representation

1. UKDA

2. Germany

3. Australia

2. Sector representation

1. Funders

2. Stakeholders in DDI Alliance

3. Disciplines from outside social sciences, e.g., health

Right now we have only North America in our group and we should try to get someone from the UKDA as the DMP Online development team looks to the UKDA for advice. There is a lot happening in Australia around metadata and they have a big investment in RDA. We can approach Steve to see if he can suggest someone. Germany is doing a lot also.

The mix of stakeholders from our community is another issue for the group. Funders are a key community we may want to approach.

We will clean up our ToR and have a second call for members of the WG. We will start our work by looking at the paper that Arofan and Tito put together. We can also target people going to Dagstuhl to represent other metadata standards around this use case of DMPs and interoperability as we should get their feedback. The mechanism of exchange is another item to focus on.  

1. Length of Operation

1. Built-in sunset clause

We will set ourselves a year to get the initial work done and review at the end of the year as to whether we have a new set of terms we should address or whether we are satisfied with what we have done. IASSIST in Bergen will be our review point.

1. Activities around Active Data Management Plans

1. Research Data Alliance Interest Group

2. Jisc DMP Ethics elements in CASRAI dictionary

3. Canadian national DMP template for data stewardship and the Portage DMP Expert Group

4. Canadian Builder and API Services

5. The DCC DMP Online development team is looking at supporting a lifecycle model for a EU Horizon 2020 project, but hadn't considered the use of an underlying metadata model.

6. DDI Lifecycle specification and overlap with ADMP elements

7. DART

We should let the chair of the ADMP RDA Group know that we exist with our revised purpose. Chuck or Mary can submit this to the chair by next Friday, September 21.

The JISC concepts around ethics are now in the CASRAI dictionary.

Chuck and Amber are the the Portage Expert Group on DMPs, so there is an opportunity to share with them. They are receptive to building from a metadata model.

Our work may influence the DCC DMP Online Tool to implement an underlying metadata model. There was an effort to reuse the administrative information in grant applications by RESHARE at the UKDA.

1. Schedule for meetings

1. Monthly; day of week; time of day

If we could articulate a use case for Dagstuhl, we should initially meet bi-weekly rather than monthly. We should try to meet in the middle of September, late September, and a few days before the Dagstuhl meeting. This time every other Friday will work.

Mary will post everything into the Google docs folder and Amber will do a gap analysis to reveal the direction we should take.

1. Other Business