Academia's buried treasure

Data created in the course of research often goes unpublished and is soon buried in departmental or even individual systems. Yet the amount of data generated is rising at such a rate that institutions have realised that they need to do something about it. Moves are afoot to tame the data deluge and harness it so that it can be an accessible and shared resource.

University librarians have had little access to research data in the past but now they need to make it their business. If data is to be a reusable and shareable resource it has to be managed, indexed and archived. Who better to do this than information professionals?

With researchers tending to see their data as their baby, librarians should start getting involved now if an unseemly custody battle is to be avoided subsequently.

The need for managing research data has been much talked about over the past five years. At last it seems that words are giving way to action, as the UK Research Data Service (UKRDS) completes a feasibility study that will recommend a model for a national data management service. Critically, it will also set out how much money will be needed to fund it all.

Research Libraries UK (previously known as CURL) and the Russell Group of IT directors got together to bid for £200,000 of HEFCE funding for the study. The also contributed some funds of their own, as has JISC.

Consultancy Serco is running the project, which has surveyed 700 researchers and focused in depth on four universities: Bristol, Oxford, Leeds and Leicester.

Precious substrate
Jean Sykes, librarian and director of IT services at the LSE and director of UKRDS, says: “A major opportunity to share and exploit research data is being lost. The data underlying a research project can be extremely important, and can be regarded as part of the research output just as much as the finished experiment or published article.

“Indeed, increasingly ­ in the social sciences and humanities in particular ­ researchers are looking for access to both the published output and the research data underpinning it, and often this is not possible because the data is not properly structured; there are no metadata descriptions, there is no easy technological access, etc.”

Sykyes points to the blurring of boundaries between disciplines; for example, health data is of interest and use to both medical and social science researchers. Yet it is not easy for researchers to share data because of a lack of standards and protocols and poor awareness of other data.

“If research data was managed throughout an agreed lifecycle and in line with agreed standards, and made accessible to other researchers, it could be reused and repurposed, thereby saving significant researcher time,” says Sykes.

UKRDS is not just out to save researchers time. The driving force behind the funding for this service is to boost UK innovation and competitiveness and ultimately the economy. Sykes believes this is a realistic aim.

“The UK is already well known to be punching above its weight internationally in terms of its research output,” she says. “The research data layer which sits under the finished research is an invisible asset which, if made accessible, could enhance and increase the total research output of this country considerably and allow the UK to increase rather than just maintain its strong global position for research excellence.”

Other countries are already tackling the issue of data management on a national scale. “It would be unfortunate if the UK did not move ahead with a UK-wide scheme,” says Sykes, “particularly given the fact that there are already many relevant building blocks in place, which is not the case in other competitor countries by any means.”

The race is on. The Australian National Data Service is already ahead of the UK as it has received significant government funding and is building its data service now.

But grand plans for a national service and high ideals about sharing data all hang on researcher willingness to play ball.

UKRDS says that although it is still working on the best model for delivering the service, the service is likely to take the form of a staffed resource offering guidance and taking into account a mixed economy. For example, where there is an existing data sharing and storage resource for a particular discipline, this will be left well alone.

“We try to take cognisance of what is going on elsewhere as there is a lot of good work going on,” says UKRDS project manager John Milner. “We wanted to avoid traps, and wheel reinvention is one of them. The sheer scale of the task is mind-bending and you need to avoid attempting to boil the ocean.”

Like this story? Spread the news by clicking below:

 del.icio.us     Digg this     reddit!