What can we learn from the research org id initiative?
In a rather confusing, but nevertheless understandable, namespaces clash, just around the time org-id.guide was surfacing to the world, the research community launched the Org-ID project to create an open dataset for recording researcher affiliations. This post considers what can be learned from this initiative.
One of the issues we’ve continued to grapple with in standards like IATI and Open Contracting relates to the identification of government entities. Unlike companies, which often have reasonably clear boundaries, official registrations, and assigned identifiers — government agencies can be much trickier to pin down, particularly when working out how to identify the recipient of aid funds, or the buyer for a contract.
Whilst the org-id.guide methodology has helped us discover a number of official lists that do include government entities, the chance of finding every single government agency around the world you want to identify in such lists is low. In the working paper published last year by the Joined Up Data initiative, we came to the view that some sort of ‘identifier of last resort’ service may be needed, to maintain a list of identifiers where no official ID has been issued. So far though, that proposal remains at idea stage — yet to move forward.
Focus on process
At PIDAPalooza in Girona this week I’ve been hearing about how very similar challenges exist in the research world: where there are well developed identifier schemes for researchers and publications — but associating these with organisations and organisational units (universities, cross-institutional research groups, departments, campuses and so-on) proves particularly challenging. Although a range of third-party issued identifier sources exist, each of the current datasets covering research institutions have weaknesses in terms of data license, governance or the extent to which the data can be easily updated.
The ‘Org ID Initiative’, kicked off by a collaborative of DataCite, ORCID, and Crossref, has run a year-long process to explore establishing a new open registry that will:
“support reliable, open, permanent, and unambiguous identification of organizations with whom researchers are affiliated.”
Alongside scoping requirements and framing principles for an open register, the working group, set up in early 2017 to take forward the initiative, has placed a strong focus on governance, process and getting stakeholders engaged.
Key to this appears to be a recognition of the politics present when lots of institutions already create, use and provide identifiers — and the need to create a sustainable solution, with an appropriate funding model that can support both data quality, and maximum openness. The working group reported back on a Request for Interest that had received input from over 20 groups expressing interest either in hosting the new open register, contributing data from existing registers, or providing technical, marketing and outreach support.
Towards a government identifier register?
The org-id.guide list of lists has been possible to bootstrap through a light-touch collaboration of standards. Building a register that actually contains details of organisations and maintains reliable meta-data about them is likely to take much more work in order to deliver a sustainable and scaleable solution.
Perhaps there is a model to pick-up from the researchers organisation identifier collaboration? When it comes to government IDs, are there actors ready to lead the call for an working group? Do we have stakeholders able to provide the start-up resourcing to move things forward?
In just a year the research community has made big strides moving from idea, to action plan. They appear to have brought together both producers and consumers of government identifiers. They have acknowledged, engaged with, and managed, the vested interests of existing providers who have been filling the identifier gap with more or less proprietary solutions — as well as articulating clear principles that should make sure any new solutions avoid capture. They have taken seriously questions of sustainability, and the kind of infrastructure that might be required to develop a long-term solution — with a number of those expressing an interest in running the open register coming from library and established institutional backgrounds.
We launched org-id.guide as a community project at the 2016 International Open Data Conference in Madrid. With IODC 2018 in the diary, taking place this September in Argentina, could we be seeing real progress on a government identifier register by then?
