Automatically creating records using ORCIDs, ORCID Public API

We have made a great deal of headway during December regarding linking up our Research Information System (RIS) and the Public ORCID API. The plan is to enable staff to 'claim' academic works that appear in their public ORCID record (see last post for details) for inclusion in our internal RIS. Since September we have enabled our internal HR system to capture ORCIDs from staff, and these are now passing into our RIS where we can start doing things like:

1. Semi-automatically populate staff output records from ORCID data
2. Track collaborations at the individual and institutional level

So far it's been relatively trivial to setup the corresponding connections between our RIS and ORCID. Unfortunately, the data coming back from the API has been a little disappointing, and in particular the structure. Some of the XML formatted information follows a schema of sorts, and some does not. Where it does it makes it easy to ingest and create associated records at our end. Where it does not we are essentially tasked with making sense of 'free text', where author, title, publication data can often be mixed up together as one string without any obvious method of delimiting the constituent parts.

I hope to chat through the issue of free text with the UK based ORCID developer at the next ORCID pilot meeting in January to see what options we have moving forward.

External Academic Repositories and ORCID

As part of Information Services and Systems' ORCID outreach programme we have been conducting interviews with academic colleagues in order to better understand there publishing workflows. Part of the interview process has related to identifying preferred open access publishing platforms by specific discipline as alternatives to large publisher or institutional repositories (such as Swansea University's Cronfa). In particular we have been talking to physicists working in the area of high energy physics at our College of Science about how they go about publicizing their research and in particular how we can encourage use of ORCID and submission into our Research Information System (RIS).

In the world of physics the challenge of Open Access Publishing appears to have been solved a long time ago with the creation of websites such as arxiv.org and INSPIRE. In a brilliant example of grass routes activity our colleagues in physics have independently contacted INSPIRE to ask if they can provide a 'connector' between their repository and ORCID to allow staff with ORCIDs to 'claim' their works published on INSPIRE without having to type the details into ORCID. The folks at INSPIRE responded very quickly and set up the connector which is live now.

This is a really good example of something quite simple that can be accomplished in a very short time that has massive benefits for a discipline such as high energy physics. With the existence of this connector our colleagues are now much more likely to surface their works in our internal RIS system as soon as we implement our own connector to the ORCID public API as discussed in last month's post. And it all happened with minimal fuss or involvement from us in IT services too! If this model could be made to work for other discipline centric online databases the life of the RIS manager would be made so much easier.

Whiskeys all round!

Accessing the ORCID public API

As part of our ongoing investigation into how we can promote ORCID use at Swansea University the Information Services and Systems Web Technology Development Team (Web Team) has been looking a machine to machine communication with the ORCID service.

The Web Team was interested to find out what can be accomplished without becoming a full ORCID member in terms of data harvesting of ORCID information relating to published works. There are 3 specific endpoints available publicly from ORCID (public API) as detailed at http://support.orcid.org/knowledgebase/articles/132271-retrieving-data-with-the-public-api. These are:

Bio	/orcid-bio	Given a contributor, return name and affiliation data.
Works	/orcid-works	Given a contributor, return the list of works he has contributed to.
Full	/orcid-profile	Given a contributor, return list of works he has contributed to, name and public affiliation data.

We are especially interested in the potential to retrieve information relating to published works and automatically creating associated records in our Research Information System (RIS) so the endpoint '/orcid-works' seemed promising. To get back records from this you need to pass in an ORCID ID so a URL such as http://pub.orcid.org/v1.1/0000-0002-9941-1955/orcid-works should work. By default sending this URL returns an XML record along the lines of:

<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
<orcid-message xmlns="http://www.orcid.org/ns/orcid">
    <message-version>1.1</message-version>
    <orcid-profile type="user">
        <orcid-identifier>
            <uri>http://orcid.org/0000-0002-9941-1955</uri>
            <path>0000-0002-9941-1955</path>
            <host>orcid.org</host>
        </orcid-identifier>
        <orcid-preferences>
            <locale>en</locale>
        </orcid-preferences>
        <orcid-history>
            <creation-method>website</creation-method>
            <submission-date>2014-04-16T08:31:09.491Z</submission-date>
            <last-modified-date>2014-07-08T13:27:31.232Z</last-modified-date>
            <claimed>true</claimed>
        </orcid-history>
        <orcid-activities>
            <orcid-works>
                <orcid-work put-code="12640508">
                    <work-title>
                        <title>Creating an in-house content management system</title>
                    </work-title>
                    <journal-title>VINE</journal-title>
                    <work-citation>
                        <work-citation-type>bibtex</work-citation-type>
                        <citation> @article{Roberts_2002, title={Creating an in-house content management system}, volume={32}, ISSN={0305-5728}, url={http://dx.doi.org/10.1108/03055720210473038}, DOI={10.1108/03055720210473038}, number={2}, journal={VINE}, publisher={Emerald}, author={Roberts, Alexander}, year={2002}, pages={30–33}}</citation>
                    </work-citation>
                    <work-type>journal-article</work-type>
                    <publication-date>
                        <year>2002</year>
                    </publication-date>
                    <work-external-identifiers>
                        <work-external-identifier>
                            <work-external-identifier-type>issn</work-external-identifier-type>
                            <work-external-identifier-id>0305-5728</work-external-identifier-id>
                        </work-external-identifier>
                        <work-external-identifier>
                            <work-external-identifier-type>doi</work-external-identifier-type>
                            <work-external-identifier-id>10.1108/03055720210473038</work-external-identifier-id>
                        </work-external-identifier>
                    </work-external-identifiers>
                    <work-contributors>
                        <contributor>
                            <credit-name visibility="public">Alexander Roberts</credit-name>
                            <contributor-attributes>
                                <contributor-role>author</contributor-role>
                            </contributor-attributes>
                        </contributor>
                    </work-contributors>
                    <work-source>
                        <uri>http://orcid.org/0000-0002-3054-1567</uri>
                        <path>0000-0002-3054-1567</path>
                        <host>orcid.org</host>
                    </work-source>
                </orcid-work>
            </orcid-works>
        </orcid-activities>
    </orcid-profile>
</orcid-message>

As detailed  at http://support.orcid.org/knowledgebase/articles/132271-retrieving-data-with-the-public-api HTML and JSON formats are also available for this data. The availability of this API endpoint is very encouraging as a means to consume records which have been made public by ORCID holders and allows us to move forward with our plan to automatically populate our RIS with ORCID information where available. Over the next few months we will work towards this functionality.

Integrating ORCID with Agresso Business World and MIS systems at Swansea University

Information Services and Systems at Swansea University has been busy integrating ORCID capture with our Human Resources system, Agresso Business World(ABW). After having moved to using ABW only recently we have successfully added the facility to allow staff to add their ORCIDs to the system, thus allowing reuse across distributed systems such as our Research Information System and staff profile pages on the University website.

ORCID Welsh Language Resources

Swansea University is delighted to announce that we have been working very closely with orcid.org to translate ORCID outreach materials into welsh. These can now be found at:

http://orcid.org/content/orcid-ambassadors-1/outreachresources

ORCID integration with RIS

Our Research Information System (RIS) has now been extended to include ORCID information for additional authors cited on research outputs. This will make it easier to track personal, geographic and institutional collaboration and interaction between staff and their peers. The ultimate goal being to facilitate data mining and visualization to better inform strategic planning and relationships moving forward. Historically information regarding collaboration and interaction has been and continues to be very challenging to collate. The hope is that ORCID will greatly assist in this effort by providing easy tracking of individual networks.

Work is now going on behind the scenes to extend our HR system to include ORCIDs, watch this space!!

JISC ORCID Kick Off Meeting Notes

Yesterday saw the formal beginning of the JISC-ARMA ORCID pilot project with a kick off meeting at the swanky Hatton – etc.venues in Hatton Garden, London. Tempted as I was to splash out on a great big diamond for my lovely wife as I wandered past the expensive jewellery shops, I managed to resist and arrive at the meeting with credit card intact.

The morning comprised introductions from Simon Kerridge from ARMA (https://www.arma.ac.uk/), the JISC Project Manager, Vereena Weigert and Laura Haak, from ORCID. Following this, every participating institution provided a brief overview of there current situation regarding ORCID takeup and plans for the future. It was interesting to note the different methods and approaches to encouraging ORCID uptake, from the very soft to the hard line.

Snippets that caught my attention included

• Portugal has a national CRIS (Current Research Information System). Of particular interest to me as I believe that such systems should be maintained and operated at the national level rather by individual institutions. Along these lines, York University shares an Institutional Repository with Leeds and Sheffield Unis under the 'White Rose' brand.
• Kent University is focussing in particular on early-career researchers and Post Grads. Using research students as ORCID champions!
• Many institutions are looking at different methods to encourage users to engage with ORCID, including mixing pre-registration with manual linking of ORCIDS with existing Intranet records
• HESA (Higher Education Statistics Agency) now includes ORCIDs in returns.

Swansea University is interested in providing ORCID bilingually to the people of Wales and with this in mind I was lucky enough to catch up with Will Simpson, a developer with ORCID, over lunch. Will explained that it's pretty straightforward to translate the various language files that feed the ORCID website, and they can be found on GitHub, which is great news! Always nice to see organisations embracing open source.

The afternoon saw everybody participating in round table discussions regarding various aspects of the pilot project. I joined a table that included folks who were interested in developing against the feature rich ORCID API (Application Programming Interface). There are loads of different ways of accessing the (public) information in ORCID and they make it easy to consume by providing it in XML and JSON formats.

I left the session and caught the train home with lots of bright ideas and a spring in my step!