Recommendations Consultation. EOSC-A Long Term Data Preservation Task Force

The EOSC Association's Long Term Data Preservation Task Force (LTDP-TF)[1], which functions under the Advisory Group (AG) Technical challenges EOSC[2] has previously shared an overview[3] document for public feedback. A paper on the frame of reference used to guide the task force approach was presented[4] at the PV2023 conference.

In this initial consultation draft of the task force outcomes the decision has been taken to minimise prose and maximise the presentation of brief assertions and recommendations for open feedback. Numerous scoping assertions are included because the work of the task force to date has revealed varying interpretations of some key research infrastructure concepts that need to be contextualised[5]. This approach supports granular feedback on each assertion and recommendation that will inform future interactions of this document. 

We would appreciate responses from a wide range of stakeholders across the digital object management lifecycle. Respondents are asked to propose amendments to text or provide explanatory comments if they object to the statements as presented. The feedback template can be found here which can be sent to preservation-tf-feedback@eosc.eu. 

The results of this consultation period will be used to revise the assertions and recommendations for future reference by a range of actors. These will inform a final report for this iteration of the LTDP-TF alongside proposals to the EOSC Association on how further progress can be guided, monitored and supported through their activities, including, but not limited to future task forces. 

The work of the task force benefits from numerous prior efforts in research data management, curation and preservation, particularly the FAIR Forever report[6], which notes that "Digital preservation involves the series of managed activities necessary to ensure the continued access to research data for as long as necessary, which encompass actions and interventions throughout the lifecycle—not just at the creation of FAIR data or the transfer and ingest to a certified archival repository." The task force's overview discussion paper[7] made it clear that the TF perspective acknowledges the importance of FAIR data and trustworthy repositories as central to the EOSC vision, but also notes that a full (meta)data lifecycle perspective is critical. The overview further highlights that to achieve preservation outcomes we depend on sustained and sustainable preservation systems that take responsibility for monitoring the technical and user environments and, where necessary, take preservation actions on digital objects (data and/or metadata). These preservation systems (including, but not limited to entities that self-describe as libraries, archives or repositories) exist within a wider network of data and metadata services. Trust across these services, through transparent practice, is vital to the success of any federated research infrastructure, including EOSC. 

While the task force notes the importance of agreeing on standards and developing metrics and tests for assessment, it is important to recognise that data services and digital objects are in the relatively early stages of a journey towards trustworthiness and FAIRness. Successes should be acknowledged and credited but less mature data services and digital objects must be invested in for improvement. The Task Force has worked to provide targeted guidance through The Strategic Research & Innovation Agenda and its Multi-Annual Roadmap[8] that have already resulted in calls that will invest in preservation. However, the journey towards agreed community criteria and broad adoption must be seen as a marathon and not a sprint. 

The scale of data and metadata collected by, created by, or of interest to researchers of all kinds, whether academic, commercial or public is vast. The baseline task of identifying and providing basic deposit, storage and access for these digital objects is already a challenge. Initial curation to ensure digital objects meet desirable criteria, such as FAIRness, requires an initial investment. For digital objects with a long term value we require the resources and skills to deliver preservation systems, actions and outcomes. Meanwhile, appraisal (review, evaluation, selection and decisions on the level of care to be applied) and reappraisal of digital objects for their value as digital assets is complex but necessary. The cost of inaction is the potential loss of resources necessary for reproducibility, replication and reuse; the benefit of action is the continued availability of resources for new and novel explorations within and across disciplines. 

References to 'digital objects' within this paper are not limited to data points created or collected during the course of research and include a wide range of data and metadata constructs, including software, that are relevant to EOSC. It may not be possible, practical or even desirable to retain, curate or preserve every digital object, but any related appraisal decisions should be transparent and the resultant levels of 'care' (retention, curation or preservation) being offered by repositories and received by digital objects should be clear. Addressing individual object types, disciplinary issues and aligning with the full contextual legal and interoperability framework of the EOSC are prerequisites, but beyond the scope of this task force and paper.

[1] https://www.eosc.eu/advisory-groups/long-term-data-preservation 

[2] https://eosc.eu/eosc-task-forces 

[3] EOSC Preservation: Overview Discussion Paper https://doi.org/10.5281/zenodo.7516259 

[4] Preservation in the context of EOSC. Sustainable repositories curating digital objects from a long-term FAIR enabling perspective https://indi.to/dpnjt 

[5] Where the TF has used an internal working definition this is flagged with WG in superscriptWG

[6] FAIR Forever? Long Term Data Preservation Roles and Responsibilities, Final Report https://doi.org/10.5281/zenodo.4574234 

[7] EOSC Preservation: Overview Discussion Paper https://doi.org/10.5281/zenodo.7516259 

[8] https://eosc.eu/sria-mar