Hervé L’Hours, Repository & Preservation, UK Data Service, UK Data Archive, University of Essex, in collaboration with the CoreTrustSeal Standards & Certification Board.
This post is a collaboration between the UK Data Service[1] and the CoreTrustSeal[2] for World Digital Preservation Day[3].
In 2024, a CoreTrustSeal board position paper defined a set of curation and preservation levels[4] that organisations can use to document how they care for digital objects. The levels range from simple storage of objects that are distributed as they were deposited, to the long-term responsibility for reuse of the data and metadata, including making changes based on the needs of the repository’s user community. The levels help define whether a repository provides active long term preservation, necessary to be in scope for CoreTrustSeal certification. However, the levels have wider relevance to defining a range of data and metadata services and responsibilities.
The paper, iterated through community feedback, resulted in four levels:
-
Level zero: content distributed as deposited. Unattended deposit-storage-access.
-
Deposit compliance: repository checks against a deposit compliance criteria.Non-compliance triggers rejection, or requires initial curation.
-
Initial curation: repository takes action as required to meet defined curation criteria.
-
Active preservation: long-term responsibility to take action as required to ensure reuse.
The levels may be cumulative as a repository progresses through them i.e. a repository that undertakes active preservation (A) is also likely to check an object meets documented criteria (D) and perform some level of curation (C).
Following the position paper the CoreTrustSeal released a related discussion paper [5]proposing a repository typology implied by the levels:
-
Retention Only (no further responsibilities)
-
Deposit Compliance (no initial curation or active preservation undertaken)
-
Initial Curation (could also offer deposit criteria, not active preservation)
-
Active Preservation (could also offer deposit criteria & initial curation)
In response to the paper, a GESIS[6] Visiting Researcher session was organised to consider the metadata implications of the levels. The result was a working paper [7] that explores the metadata characteristics of digital objects and repositories that support the curation and preservation levels, including retention information and reappraisal triggers.
D. Deposit compliance
Deposit compliance checks allow a repository to focus on digital objects relevant to their users, and ensures data owners know whether their data is in scope. They include criteria for acceptance (e.g. disciplinary scope or minimum metadata), and potential outcomes. For example, the UK Data Service may[8]:
-
Accept into the main repository collection.
-
Accept into the self-deposit ReShare[9]
-
Advise an alternate place of deposit.
C. Initial curation
Initial steps to prepare and enhance an object for storage, sharing and re-use include checks, corrections and enhancements to the metadata, documentation and even the data, and possibly conversion to new formats. It should be clear what steps are taken and what standards achieved[10]. Initial curation may also include preparations for active preservation if that is part of the repository mission.
A. Active preservation
Active preservation are steps that a repository takes to ensure that data and metadata are accessible to and usable by its community for the long-term. This includes conversion to newer formats to meet changing user needs or in the face of technical obsolescence. At a minimum, metadata at the repository level should state any preservation periods and any potential triggers for preservation actions (e.g. updated metadata schema).
Other metadata characteristics
Retention of digital objects is the foundation of all levels of curation and preservation and structured metadata, and retention periods should form part of structured metadata. No level of retention or care can be guaranteed into infinity as users’ needs, digital objects’ values and repository remits may change, so clear reappraisal periods and criteria are necessary.
Defining and implementing additional metadata around levels of care provides broad community benefits. It would strengthen links between repositories and object information, and demonstrate greater provenance, transparency and interoperability, increasing data owners' and users' trust. It would also allow for greater automation, including alerts and triggers for reappraisal and preservation actions. Third parties, including registries, would be able to harvest richer information about repositories and objects.
Future iterations of this work will be delivered by CoreTrustSeal and others, including the FAIR-IMPACT[11] project. If you would like to share your thoughts or disseminate them both the discussion paper and the working paper are open for public comment. We think defining curation and preservation levels this way has real potential for the wider community, as well as being a step towards more streamlined and automated assessments.
[1] https://blog.ukdataservice.ac.uk/
[2]https://www.coretrustseal.org/
[3] https://www.dpconline.org/events/world-digital-preservation-day
[4] CoreTrustSeal Standards and Certification Board. (2024). Curation & Preservation Levels: CoreTrustSeal Position Paper. Zenodo. https://doi.org/10.5281/zenodo.11476980
[5] CoreTrustSeal Standards & Certification Board. (2024). Types of Repository: Entities, Responsibilities, Objects. CoreTrustSeal Board Discussion Paper (v01.00). Zenodo. https://doi.org/10.5281/zenodo.13133041
[7] L'Hours, H., Kleemola, M., & Recker, J. (2024). CoreTrustSeal Levels of Curation and Preservation: Implied Repository and Object Metadata Characteristics (v01.00). Zenodo. https://doi.org/10.5281/zenodo.12701324
[8] https://ukdataservice.ac.uk/help/deposit-data/deposit-in-the-curated-data-repository/
[9] https://reshare.ukdataservice.ac.uk/
[10] https://www.data-archive.ac.uk/managing-data/digital-curation-and-data-publishing/quality-control/