THOR Ambassador Newsletter 2017-06

Press/Media: Other

Description

Let’s talk to…

 
Stian Soiland-Reyesorcid.org/0000-0001-9842-9718

Stian is a Technical Architect and Research Associate at the eScience Lab at the School of Computer Science, University of Manchester.

Why did you decide to become a THOR ambassador?
I've supported ORCID since the early days, and worked closely with people in the THOR project, in particular when I added the Linked Data support to the open source ORCID code base at a shared hackathon - see for example content-negotiation on orcid.org/0000-0001-9842-9718 which redirects here with a valid FOAF/WebID profile as well as provenance information about the record itself.
What I perhaps should have worked more on is promoting ORCID further within the Linked Data community, add documentation and expand the ORCID code base to use newer Linked Data vocabularies such as http://schema.org/.

Can you tell more about the way you are promoting PID use within your organisation/community?
As I am working in the Linked Data community with a focus on describing provenance and metadata for open research, I have found permanent identifiers such as ORCID for identifying people are essential. 
For instance in the W3C HCLS profile we recommend ORCID to identify individuals, and similarly recommend ORCID in the provenance information in our Research Object specification for sharing reproducible research data and methods.
While the Linked Data community have long been promoting a distributed system with WebIDs - it's easy to make your own URIs - most researchers need a simpler solution than hand-editing RDF or maintaining their own servers. I also think longevity is important; "under the desk" solutions tend to die after a few years.

ORCID has already got big traction and support, as exemplified with the 685 member organisations and in my opinion ORCID covers the 95% use case for researchers. I wish however there were more work on ORCID integration - it doesn't have to be fully fledged single-sign on integration with ORCID, just a "ORCID (recommended)" field in addition to the traditional "email" and "home page" fields (which vary over time).  Any repository can do this, no need for ORCID membership. Just add a simple field with some simple messaging to ensure a uniform format of the recorded ORCID identifiers (e.g. with or without "http://" and "orcid.org"). Actually I have found myself "cheating" if a submission system like EasyChair doesn't support ORCID iDs: I fill in my ORCID URL as the "Web page" - even though it is not true - ORCID identifies me as a person, and I am not a web page!  See:https://www.w3.org/TR/cooluris/#semweb.
So I bite my semantic web tongue together and use this workaround until ORCID is integrated properly in systems we academics use. But this only works for idealists like me - regular authors need to be asked straight up "Do you have an ORCID iD?".
I hope projects like THOR can help integrate ORCID more with such systems and repositories, perhaps provide code patches, so that they can ask for ORCID iDs in a friendly and systematic manner, as well as presenting this nicely for any resources.

Can you explain more about the FAIRDOMHub and your role in this initiative?
The FAIRDOMHub is an open web-based repository for sharing heterogeneous scientific research datasets, models or simulations, processes and research outcomes, in particular within Systems Biology. The hub is part of theFAIRDOM project, which helps researchers to make their data and models adhere to the FAIR principles: Findable, Accessible, Interoperable and Reusable.
We ask about the ORCID iD as part of the FAIRDOMHub user profile, and then we can use it in the Linked Data view on the metadata from the hub to uniquely identify researchers across their institutions and projects.
We did actually receive some negative comments about asking for ORCID iDs, particularly from German researchers with privacy concerns, as they noticed that ORCID's privacy policy means any arbitration have to be done not just outside EU, but in person in New York, and they would have to pay half the cost no matter the outcome. I agree that these are valid concerns, and people should be able to choose for themselves when and if to apply or use their ORCID iD - although the identifier is public it should be considered sensitive, just like anything else that directly identifies a person. So we now made ORCID optional in the FAIRDOMHub profile. I also notice that there's a new ORCID DE project which hopefully will address some of these privacy concerns.

Would you like to add anything else?
In BioExcel, a Centre of Excellence for computational biomolecular research, BioExcel we are making a cloud portal to browse and execute BioExcel tools on multiple cloud providers. We are using the ELIXIR Identity service to do authentication, which supports both institutional login and ORCID single sign on.
As The University of Manchester is also an ORCID member and I can log in to ORCID using my institutional login from the University, perhaps a bit absurdly this means I can log in to our BioExcel cloud portal using my ELIXIR iD, where I log in with ORCID, where I log in with my University SSO, and magically it all redirects back again correctly to our portal. If I choose to log in to ELIXIR with Manchester SSO instead, I am still the same user, as the ELIXIR Identity service has linked the different accounts. Since it's seldom I am logged out of all these services, it is usually quite easy! Single Sign-In that works!

Thank you!

Period13 Jun 2017

Media contributions

1

Media contributions

Research Beacons, Institutes and Platforms

  • Institute for Data Science and AI

Keywords

  • orcid
  • linked data
  • fairdom