I attended the Research Information Network’s event, “How do we make the case for research data centres?“. The event was to mark the launch of the RIN/ JISC report, “Data centres: their use, value and impact“. It was pretty high-level stuff, with plenty of discussion about the relationship of data centres to the research process and ways in which datasets are curated. There were a couple of very interesting examples of use of data are used by researchers themselves, one from an academic who noted the value of data centre data because it didn’t require awkward conversations with potentially rivalrous labs; and another from a researcher in a small company building socio-economic models using data derived from ESDS’ wealth of datasets.
There were a few lessons for City Research Online, though, and I outline them briefly here:
- Institutions (and by extension institutional repositories) remain important for the curation of data, given their local knowledge and relationships with researchers.
- Institutional repositories are an excellent and cost effective method of storing data. What they are less good at is the managerial aspects of serving datasets to those whom might wish to access them. IRs can’t provide the rich metadata and sophisticated web front ends that dedicated data centres provide.
- However, there is still a role for data curation by IRs, for simple and/ or small datasets. Where data are presented in tabular form and are easily catalogued, IRs can take on data that would not be interesting for data centres.
- This is my own opinion, but I would extend the above role for IRs to include datasets which underlie published journal articles, particularly in those cases where we already archive the paper(s) in question- the ability of IRs to link together items is of benefit here. The challenge here is to advertise this as a viable and meaningful service for data creators.
So, some challenges for us to look at. In my experience, datasets are one of those repository things that can be “worried about later”, but I also think that datasets are of increasing importance to research. If we can identify ways in which City Research Online can usefully provide (perhaps modest) data curation services, then so much the better.