Draft case statement for ADMP Interest Group

20 Oct 2014

Comments please by COB Monday 27th Oct on the following: after we'll put the case systement into the system to seek approval.

 

Draft Charter for Active Data Management Plans (ADMP) Interest Group

Co-Chairs: David Giaretta (UK), Helen Glaves (BGS, UK),  David Baker (CASRAI, USA)

 

The proposed activity of this group is to act as a nucleus for discussing requirements for and identifying developments needed to support active (i.e. able to evolve and be monitored) data management planning. Working groups will be proposed to carry out work on specific areas of interest.

Currently research data management plans (DMP), created at the proposal stage of a project, do not evolve and cannot be monitored in any detail. The DMP should be the first step in the data life cycle for any dataset and therefore fundamental to ensuring that data is appropriately archived, preserved and available for re-use.   

Initial meetings have demonstrated that there is very broad interest in this topic and also a willingness to contribute to the work. Preliminary ideas around the concept of ADMPs and  potential topics for the working groups have been identified. These include specifying practical tools and services to support  the data creators in making their data re-usable and also as the data managers and funders that have a requirement to administer and monitor these plans.  The ADMP IG is distinct from, but will need to work with, many other groups including those on certification, domain repositories, metadata and preservation e-infrastructures.

  • Esther Conway's picture

    Author: Esther Conway

    Date: 20 Oct, 2014

    Hi just a couple of comments here, regarding the scope of the proposed ADMP, from the CEDA  (British Atmospheric Data Centre and  National Earth Observation Data Centre) archive perspective. 

    In addition to dealing with data arriving from project where we have the ability to negotiate and advise data creators, we also deal with the historic archive (much of which is at preservation risk) , rescue & recovery operations in addtion to  deposit request from projects already at completion.  All of these archival holdings would benefit from dynamic data management, which could  use active data management plans.  Ideally the first step in data lifecycle would initiate a DMP but our initial engaement with many data sets occurs after this point and an ADMP must be initiated then.

    Based on models developed at CEDA for "Active Data Management" more than just "plans" need to be monitored by data managers .  Once data creator have  handed over to archives  they must monitor the preservation state  of archival holdings in relation to monitored risk, funding, skill levels as wellas fuffilemt of plans  .  I would therefore advocate the scope of the ADMP be extended to include these areas.  I attach diagram to illustrate the areaas we need to monitior in CEDA to achive "Active Data  Management"

    Unfortnately I was unable to make the Amsterdam RDA meeting s, o apologies if these issues have been discussed and are covered by  "preliminary ideas around the concept" but it is critically important for us that active managemnt plans intgrate with manageable archival processes, not simply point to tools and services.

  • Rob Hooft's picture

    Author: Rob Hooft

    Date: 20 Oct, 2014

    David,
    I think the ADMP initiative could benefit if it included not only a focus on archiving, but also include proper data management during a project. In my opinion a data management plan also should include ideas on how data will be handled during the project: e.g. how backups will be able to deal with hardware failure or human error, but also aspects as simple as the question whether all x TB of data will arrive at once, or over the course of a year can make a huge difference in both project logistics as well as project costs and should be part of the (A)DMP.
    One of my rules of thumb is that good data management should benefit the scientist performing the project, it should not be an obligation that only helps others. Better traceability of data and ICT staff that is not going to be shocked by surprises are definite benefits to a project.
    Rob

  • Reagan Moore's picture

    Author: Reagan Moore

    Date: 20 Oct, 2014

    David:
    The Practical Policy working group welcomes examples of policies / procedures that actively enforce management policies, automate administrative functions, and validate assessment criteria. We seek implementations that include:
    - English language description of the policy
    - actionable implementation of the associated procedure. This requires choosing a technology framework (iRODS, GPFS, DataVerse, etc.)
    Will your interest group identify procedures / workflows that can be applied within some technology framework?
    Thanks,
    Reagan Moore
    From: dgiaretta <***@***.***>
    Reply-To: "***@***.***-groups.org" <***@***.***-groups.org>
    Date: Monday, October 20, 2014 at 7:07 AM
    To: Active Data Management Plans <***@***.***-groups.org>
    Subject: [rda-datamanagplans] Draft cae statement for ADMP Interest Group
    Comments please by COB Monday 27th Oct on the following:
    Draft Charter for Active Data Management Plans (ADMP) Interest Group
    Co-Chairs: David Giaretta (UK), Helen Glaves (BGS, UK), David Baker (CASRAI, USA)
    The proposed activity of this group is to act as a nucleus for discussing requirements for and identifying developments needed to support active (i.e. able to evolve and be monitored) data management planning. Working groups will be proposed to carry out work on specific areas of interest.
    Currently research data management plans (DMP), created at the proposal stage of a project, do not evolve and cannot be monitored in any detail. The DMP should be the first step in the data life cycle for any dataset and therefore fundamental to ensuring that data is appropriately archived, preserved and available for re-use.
    Initial meetings have demonstrated that there is very broad interest in this topic and also a willingness to contribute to the work. Preliminary ideas around the concept of ADMPs and potential topics for the working groups have been identified. These include specifying practical tools and services to support the data creators in making their data re-usable and also as the data managers and funders that have a requirement to administer and monitor these plans. The ADMP IG is distinct from, but will need to work with, many other groups including those on certification, domain repositories, metadata and preservation e-infrastructures.
    --
    Full post: https://rd-alliance.org/group/active-data-management-plans/post/draft-ca...
    Manage my subscriptions: https://rd-alliance.org/mailinglist
    Stop emails for this post: https://rd-alliance.org/mailinglist/unsubscribe/45778

  • Esther Conway's picture

    Author: Esther Conway

    Date: 21 Oct, 2014

    Hi Sarah/All,

    Yes  CEDA (as part of the NERC data centre ) use data management plan for project funded from NERC grants which we agree as you said with the PI.  The big issue for us is that  the data lifcycle extends long past the end of data producing project For example CRU :

    http://badc.nerc.ac.uk/view/badc.nerc.ac.uk__ATOM__dataent_1256223773328276

    which has time series data dating back to 1901 still requires active management and planning to serve the 1000's of active researchers using this data set who had nothing .  For example we expect this data to used in 50 years time and the tools we recommend  and link to for users e.g. CDAT will be replaced by something else.  Today we were dealing with GOME data (ozone data from satellite intruments) where the scientifc user community needed us to identify and document version (according to processessing algorithm) , this was driven by requests from the active research community . Archive data management  is not simply a matter of storage and fixity checking. We would actually consider data centrel activities to be very broad and adress the needs of researchers (data producing and consuming ) and funders.  Just to answers Reagan's point quickly  this could include for us  " procedures / workflows that can be applied within some technology framework".

    We would consider data management plans in their current form only to address the earlier phases of the data life cycle and only target data aquisitions from NERC projects. 

    While I agree completely with point made by Rob on data mangement needing to serve the project scientist, we also need to make sure it extends past the initial project.

    I hope that clarifies things a litte.  As I thing scoping out the intended purpose of "Active Data Management Plans " is an important thing to do 

    Esther

  • Sarah Jones's picture

    Author: Sarah Jones

    Date: 22 Oct, 2014

    Just re-posting my earlier message so it displays under the list of comments. I'd replied to the list which started a new thread, so just want to bring this back into the chain for clarity. This provides context for Ester's reply above.

    Hi David, all,

    I agree that the more we can make DMPs 'active' so they're updated and monitored the better. There are a couple of examples from funders where they expect DMPs to evolve. See for example NERC, which has an outline DMP at the grant proposal stage and a full plan post-award that is created by the data centres together with the PI. The Horizon 2020 open data pilot meanwhile asks for DMPs as a project deliverable that should evolve to gain more precision and substance during the lifespan of the project.
    http://www.nerc.ac.uk/research/sites/data/dmp
    http://ec.europa.eu/research/participants/data/ref/h2020/grants_manual/h...

    Rob makes an important point that the focus should be a bit broader than archiving. If the DMP is to be active throughout the whole data lifecycle, then it'll play an important role for researchers, their institutions, data archives, and funders. The ADMP group should address all of these stakeholders and their needs.

    Monitoring is definitely an area where more work could be done too, especially where aspects could be automated e.g. confirming deposit in the stated repository. I expect there's scope to build some of this into existing tools like DMPonline and DMPTool.

    All best

    Sarah

submit a comment