JISC

Science is being transformed by accelerating change in information technology, with huge increases in computing power and network bandwidth, accompanied by an explosion in data volumes and information.

Executive summary

The term e-Science – or more inclusively “e-Research” - has been used recently to describe the research culture and opportunities enabled by these developments, and the collaborations of people and of shared resources that are needed to resolve new research challenges, whether in the sciences, social sciences or humanities.. e-Science enables a new order of collaborative, more inter-disciplinary research, based on shared research expertise, instruments and computing resources, and crucially increasing access to collections of primary research data and information - the knowledge base of research. The term e-Science is applied to these techniques when applied to the sciences. In this report we use the term e-Science.

There are challenges, however: these same technology changes put the very data they create and use at risk, and raise serious and complex issues of strategy and policy regarding its creation, management, and long-term care – its curation – for which top-level responsibility urgently needs to be adopted to protect and further UK research.

Our study examined the current provision and future needs of curation of primary research data in the UK, particularly within the e-Science context. At a strategic level we found: 

• Confirmation that the data revolution presents significant challenges and opportunities. However, our surveys show that the UK is not fully prepared to capitalize on the opportunities and urgently needs to address this
• There is a lack of a government-level, overall strategy for data stewardship and data infrastructure to which research administrators can refer, still less to support researchers in their evolving roles and duties with regard to data curation
• Existing data centres are usually supported by sponsors whose primary funding focus is research projects
• The current short-term funding models for the provision of curation are antithetical to its long-term nature and needs
• There will be an exponential increase in data volumes from e-Science over the next decade as planned new scientific instruments and experiments come on stream. However, to benefit fully from this major investment, further action is needed to support the curation of the data that will be generated
• Not all primary research data needs to be retained or has long-term value. Its potential value for generating new research will vary, and the level of investment in the curation of datasets therefore needs to be identified and graduated accordingly.

At a policy level we found: 

• Provision of curation is patchy, and more advanced in some disciplines than others; the basic life sciences, and “big” collaborative sciences such as particle physics and astronomy are examples where provision is most advanced
• Where retention and curation of primary research data is a requirement set by funding bodies, the majority of researchers surveyed stated this requirement was not funded. Where guidance is provided, researchers frequently felt that it was out of date or inadequate
• Awareness of the issues - particularly data longevity difficulties - is generally low among researchers. Consequently the good practice needed to assure data longevity is rare, putting valuable resources at risk
• For curation to be effective the researcher needs to be engaged in the curation of his or her own data, working in partnership with curators. But few incentives or procedures are in place to ensure that this engagement is achieved
• Whilst practice and experience in curation is increasing rapidly, areas of curation are still in a research and proof-of-concept phase. Much research and practical, exploratory activity is being undertaken in the UK, and its quality is world-class
• The data revolution raises many issues of trust which must be addressed before data-based research can flourish – issues of security, confidentiality, ownership, assured provenance, authenticity, and data and metadata quality
• There is little interaction and sharing in curation experiences between science-based industry and the academic sector. Within the next decade the curation of digital content and data is likely to be critical to science- and engineering-based industries and to knowledge-based economic activity.

Based on these findings we set out our major, strategic recommendations in the list overleaf. Our report details further specific recommendations within the body of the report, where we also outline proposals for the organisational structuring of curation provision and provide a table showing which recommendations address the findings summarised above. These recommendations cross organisational boundaries and span organisational levels. 

Strategic recommendations

 
  
It is our view that, as the highest priority, responsibilities should be assigned for the strategic recommendations. Following feedback from JCSR, the following responsibilities for taking actions are recommended:

Action to be taken by:

Read the final report below