5

More and more research projects are now required to have a data management plan, sometimes already upon submission of the proposal, sometimes early into the then funded project.

These plans lay out the kinds and amounts of data to be generated by the project, as well as information about their storage, archiving, accessibility and related questions. There are also software management plans or more generic project management plans, which serve similar purposes.

A big issue with these plans — which so far are usually just text documents, and often not even public ones — is monitoring compliance, i.e. if statements made in the plan have been followed up by appropriate action.

I think this problem could be eased substantially if these plans were both public and machine readable, which would also allow them to be used for things other than monitoring compliance, e.g. as a discovery tool — imagine you could subscribe to the feed of all data of kind X acquired from sources of kind Y.

What I mean by machine readability is expressed quite well in the introductory part of the Wikipedia article "Machine-readable data":

"Machine-readable data is data (or metadata) which is in a format that can be understood by a computer. There are two types; human-readable data that is marked up so that it can also be read by machines (examples; microformats, RDFa) or data file formats intended principally for processing by machines (RDF, XML, JSON).

Machine readable is not synonymous with digitally accessible. A digitally accessible document may be online, making it easier for a human to access it via a computer, but unless the relevant data is available in a machine readable format, it will be much harder to use the computer to extract, transform and process that data."

Any advice on how best to go about achieving machine readability for data management plans?

I am aware of https://dmponline.dcc.ac.uk/ and https://dmp.cdlib.org/ but as far as I can tell, their machine readability is limited to non-existant.

  • 1
    If you haven't already, I'd recommend getting in contact directly with the Digital Curation Centre as they're actively developing DMPonline and I think this is one of the areas they're interested in exploring. – Jez Feb 23 '15 at 10:57
  • Thanks. I am in contact with them, and they are interested, but it will take some time, and I've posted it here in part to get in contact with others interested in the matter. – Daniel Mietchen Mar 8 '15 at 0:56
  • 3
    I'm voting to close this question as off-topic because it's not about academia: it's a computer-science question about how to make a particular category of information machine-readable. – EnergyNumbers Aug 21 '15 at 2:59
  • Thanks for explaining your vote. The question may appear to be about computer science, but just imagine we had machine readable data management plans - how would that affect academia? For one, compliance with policies could be monitored, while on the other hand, we could use the information in the plans to build new data discovery tools. For instance, you could be notified automatically every time a dataset of a specific kind is announced in a plan or deposited in a relevant repository. We're not there yet, but I think it's relevant to academia and the life of researchers. – Daniel Mietchen Aug 21 '15 at 3:13
  • 1
    @Wrzlprmft What I mean is expressed quite well in the introductory part of the Wikipedia article "Machine-readable data": "Machine-readable data is data (or metadata) which is in a format that can be understood by a computer. There are two types; human-readable data that is marked up so that it can also be read by machines (examples; microformats, RDFa) or data file formats intended principally for processing by machines (RDF, XML, JSON)." – Daniel Mietchen Aug 21 '15 at 23:18
1

I think that one of the optimal solutions to the problem of "achieving machine readability for data management plans" could be as simple as specifying procedures of data conversion from source formats (human-readable or semi-machine-readable) into desired machine-readable formats for further data analysis. Those procedures might be coded in natural language instructions or, alternatively and preferably, via algorithms for benefits of automation and reproducible research.

Conversion from machine-readable formats to human-readable is an optional step, based on projects' requirements (while optional, this is a typical step for most projects, implemented in their reporting phase). The types and details of data conversion functionality IMHO obviously depend on projects' specifics and, therefore, should be determined and specified on a project by project basis.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.