Refining the brightest and best new ideas in research data management

Daniela Duca, senior co-design manager at Jisc, says that the co-design process’s latest initiative Research Data Spring will help the research community find solutions to some of its biggest current data management challenges

The media storm that erupted around 'Climategate' just a handful of years ago illustrates very clearly how a tendency towards secrecy can hamper research, breed suspicion and hinder progress.

Since then, there have been many developments that encourage open research practices – some regulatory, some cultural and many technological. But even now, when openness is widely accepted as a force for good, it brings challenges as well as opportunities. Most of us know a researcher who still has the instinct to keep work close to their chest and who worries about what will happen if they share data openly. And, as governments and research funders continue to apply fresh pressures on researchers and institutions to make sure research data is both discoverable and reusable, the research community needs strong, workable solutions to help it overcome those challenges.

This is one of the themes that emerged from research at risk, a co-design challenge that focuses on developing practical solutions to the big issues that research institutions face in defining and implementing effective research data management (RDM) policies, processes and systems. Ultimately, we aim to streamline some of the processes so that it is easier for universities, colleges and researchers to comply with government and funder requirements and also to remove cost inefficiencies involved in doing so.

Through research at risk we are identifying common issues so that we can share ideas and best practice and also remove wasteful duplication of development effort. Most recently, we’ve launched research data spring, a bid to create innovative partnerships between all stakeholder groups in the research process (including researchers, librarians, publishers and developers) to develop better tools, software and services to support researchers’ workflows and their management and use of research data. Through research data spring Jisc will fund members of the research community to work collaboratively to develop these tools.

So several months ago we invited researchers to contribute their best ideas for creating some practical solutions for further development – and contribute they did! Seventy ideas came in, from which we shortlisted 44 for further consideration at a workshop at the end of February. Following discussions and additional development work some of these ideas overlapped and could be meshed together so that we were left with 27 really strong ideas addressing a wide variety of RDM issues.

To give you a flavour of the sheer range of creative solutions, I’ve picked three very different ideas that were explored at the workshop:

A consortial approach to building an integrated RDM system is the concept that emerged following a meeting between five groups, each with their own proposal for systems integration. Digital data archiving service provider Arkivum, information security services provider CREST, the University for the Creative Arts (UCA), Leeds Trinity University and the University of London Computer Centre (ULCC) met in the sandpit on day one of the workshop and combined their ideas with the aim of developing a complete package of tools and services that work together to help in managing the complete RDM lifecycle, and to have connectors to local infrastructure and cloud providers. They say it could simplify RDM for institutions and allow consortia or service providers to create shared services and hosted solutions. They aim to show that it is possible to support all aspects of RDM and to give institutions a flexible solution that will enable them to select just the bits of the solution that they need. It has been fascinating to see how a group of smaller, specialist institutions can come together to create the solutions that meet their own particular RDM requirements.

Dissecting digital humanities data with biomedical tools is an idea that proposes to adapt and implement the open source DataSHIELD tool, originally developed to analyse biomedical data, and apply it to the analysis of data in the digital humanities. The co-owners  – including the University of Bristol, the British Library and the Ontario Institute for Cancer Research – point out that this would help researchers in the humanities to work around the obstacles that limit or even prevent open analysis of sensitive data. The confidential nature of personal data, licensing issues surrounding raw data and the sheer size of some datasets can all mitigate against open analysis but this concept would enable researchers to access data remotely, sending their query for analysis to the place where the data is stored. The collaborators aim to develop their proof of concept for a range of analyses using digitised books held by the British Library.

And finally, 3d data workflow. A multidisciplinary team from Bangor University has had the idea to develop an application and repository to support the growing amount of 3d data that is now being produced and to enable search, analytics, visualisation and comparison of datasets. They selected research portal HeritageTogether to develop their proof of concept. It is a place where users can upload their photos of heritage sites to be turned into 3d models, and they hope to provide a way for researchers to compare their data, make models and run simulations to test hypotheses, analyse how sites deteriorate over time and foster better understanding of heritage sites.  The team says that the idea has applications in many different disciplines, including for oceanographers modelling flooding, and medical researchers handling volumetric patient data.

The expert panel has now reviewed the 27 proposals and (by the time this story appears) they will have selected the ones that will receive initial development funding. And that’s when the hard work really starts. Over the next three months the successful project teams will work on their first set of deliverables, bringing news of their progress to another workshop in June for discussion and feedback, to decide what the next priorities are and to pitch for further funding.

This highly collaborative approach is one of the hallmarks of the co-design innovation model. By working together on common problems and reviewing them together regularly it is possible to identify what looks set to yield important benefits then commit more resources if necessary. Equally, it enables contributors to spot aspects of their project that would benefit from more thought before moving further ahead. That means that development resources are targeted to where they can bring most benefit most rapidly to the sector.

Following the summer workshop more funding will be released for a further four months of development work before another workshop in the autumn, then still more development work which, we hope, will enable the projects to develop their prototypes ready for a final showcase event in summer 2016.

To find out which projects have been selected for funding, and for updates on progress between now and next year’s showcase, please visit the research data spring and research and development pages on the Jisc website.