Last week, Elizabeth Newbold and I represented the British Library’s Science, Technology and Medicine team at the International Digital Curation Conference in Amsterdam. We bumped into a few of our previous workshop friends there, but for those who couldn’t make it, here are some things that might be of interest to you.
You tend to hear a fair few metaphors at conferences, but I saw one in a new light thanks to one of our DataCite users, Scott Edmunds from the BGI: ‘Data is the new oil’. This comparison is usually intended in the money-making sense, but my interpretation was very different: being able to get the data out of the ‘well’ is one thing, but you need to manage and curate (refine) it before it’s actually useful. If you try and put ‘crude’ data into your research, it’s going to break down. Taking that even further, I now like to think of persistent identifiers pointing researchers to where they can fill up their tanks!
Metaphor suitably stretched, we were actually there to talk about the data discovery work that we introduced at the second data citation workshop. We used our existing search tool Explore the British Library and put information about online datasets into it, providing a way for our users to find relevant data. This related well to one of the notable themes of the conference: developing data services on top of existing infrastructure. The take-home message was that many of the elements necessary for effective data management already exist – building your service is just a matter of putting together the pieces of the jigsaw.
Finally, I’m going to mention the session on data confidentiality and openness, which is directly relevant to our previous workshop on sensitive data. We heard from Louise Corti (UK Data Archive, another of our DataCite clients), Carl Lagoze (University of Michigan) and Jared Lyle (ICPSR). Jared highlighted that in sharing confidential data, you need to have Safe Data (removing personal data where possible); Safe Places (secure storage and access) and Safe People (ensuring those working with and using the data maintain confidentiality). The UKDA also uses these principles, and it was a revelation to hear how much time they spend ensuring ‘safe data’: poring over all qualitative data (e.g. reading interview transcripts) to check it isn’t disclosive. This is particularly important as the UKDA will bear the brunt of any legal action that arises should any confidential information become public.
Carl Lagoze showed how Michigan has written metadata ‘cloaking’ into their XML, which codes in the level of access required to view each metadata element, and then keeps that information behind the firewall if users are not registered with the appropriate access level. This approach may be very useful to have behind DOI landing pages, where some parts of the metadata may need to be restricted.
Posted by Rachael Kotarski.