Workbook on Digital Private Papers > Administrative and preservation metadata > Rights metadata for personal archives

Rights metadata for personal archives

Rights metadata and personal digital archives

In the context of personal digital archives, it is important to include a rights clause in all donation and deposit agreements, in which explicit permission is granted by the donor or depositor for the repository to undertake preservation and access activities. This means that at least the material under the copyright of the donor or depositor is covered by a formal permissions agreement, and the conditions of this agreement should be recorded in the rights metadata which accompanies the collection and its component parts. Copyright status metadata should be created for the rest of the material in the archive.

It is important to extract as much IPR information about a digital archive as possible at an early stage in the relationship between donor and repository. Curators should take the opportunity to discuss rights with the donor or depositor and perhaps to contact, or at least identify, other significant rights holders in the material. More detailed and accurate rights information can be obtained at this stage than if the creation of rights metadata is left until later in the record's lifecycle. All of the rights information obtained should be recorded as part of the AIP.

Below are some suggested approaches to recording rights metadata about the kinds of digital objects found in a personal archive.

Rights metadata at AIP stage

Rights information which has a bearing on preservation activities should be recorded as comprehensively as possible in the AIP metadata record. Full rights information regarding access and use should also be recorded as part of the administrative metadata; elements of this can be extracted for researcher use as part of the DIP at a later stage. Rights information need not be included as part of the descriptive metadata of an object's AIP since these are held in a stand-alone preservation environment and are not accessible to researchers.

The METS document for each digital object should include a <rightsMD> element within the Administrative Metadata Section (<amdSec>). This will be used to record IPR information relating to access and use of the digital object by researchers. The METSRights schema can record this information and should indicate whether access and use of the digital object is to be provided under statutory provisions ('copyrighted') or under contractual conditions ('licensed', 'contractual'); or whether the object is in the 'public domain'. Most digital objects will not be in the public domain for many years, and most of the material in a digital archive will not be subject to a licence, so digital archivists are likely to be basing access conditions on copyright law.

METSRights should also be used to record the full name and contact details of the rights holder or their legal representative, where this information can be ascertained. The <RightsHolderComments> element might be used to indicate where rights holder information was unavailable or could not be traced; in many cases digital archivists will not have the time and resources to trace every single rights holder represented in an archive when that information cannot be extracted from the archive itself. Although it includes general elements like <RightsDeclaration> (to contain a broad declaration of rights associated with a digital asset), the METSRights schema does not include specific elements in which to record the kind of details about copyright status recommended by Coyle, e.g. the name and dates of the author as well as the rights holder if there is a distinction, and the date of creation of the object. The copyrightMD schema could be used to record such information. Archivists should also monitor the work which is currently underway to see whether METSRights elements can be represented as an application profile of ODRL; this might enable users to enhance the METSRights record with more granular copyright information.

The <context> element of METSRights uses attributes to express the types of user to which particular permissions or constraints relate. The OTHER attribute provides a means to define local categories, e.g. Registered Readers; Remote users who have filled in an online copyright declaration form, etc. Specific types of permitted act can be defined, e.g. discover, display and copy. See Chapter 09 Legal issues for the access conditions that are likely to apply to personal digital archives which are still in copyright.

In cases where the creator of a digital object has attached a Creative Commons licence to their work, this should also be indicated in METSRights. In some cases creators might use a Creative Commons licence to dedicate their work to the public domain, in which case there will be no restrictions on its re-use. In other cases a creator will retain some rights over the material and any specific permissions and constraints should be recorded using the <Context> element of METSRights; these can be recorded using Boolean values, and a fuller explanation of these might be recorded in the <ConstraintDescription> element.

In 2005 a specification for an ODRL Creative Commons Profile was produced. This describes how the semantics of Creative Commons licences can be represented using a Profile of the ODRL Rights Expression Language; ODRL is XML-based, so this information might be extracted and included in an AIP METS file. This could be done by adding an additional <rightsMD> section, using the <mdWrap> element, and the OTHER attribute to denote that XML metadata which does not come from one of the METS extension schemas is being used.

The digital object METS document will be linked by an identifier to a related METS document containing the information which forms the PREMIS Rights entity for that object; this records rights associated with preservation activities rather than researcher access. Storing this information separately means that all the digital objects in an archive which are under the copyright of the same person and subject to the same conditions can refer to a single rights record; an example of a complete rights entity appears earlier in this chapter. While the rights entity is not a mandatory element of PREMIS, Paradigm recommends that it is used.

Currently the PREMIS rights entity only allows the recording of a limited number of specified actions (based on a controlled vocabulary) that are covered by explicit conditions. These would be agreed as part of a donation or deposit agreement, and so are only likely to apply to material in the copyright of the archive's principal creator. Currently UK copyright law is ambiguous on the subject of copying digital material for preservation purposes, and it is not clear whether digital archivists can rely on legislation as a mandate for preservation copying. However, the Gowers Review may alter this situation, and in future archivists may be able to claim the right to copy for preservation purposes under copyright law. Should this come to pass, it would be very helpful if Karen Coyle's suggested amendments to the PREMIS Rights entity are adopted and added to the Rights schema. The inclusion of further metadata elements from the copyrightMD schema or Cedars (as recommended by Coyle) would provide an even more granular record.

In the same way that digital archivists need to carry out a 'technology watch' while objects are in archival storage, it is also important to maintain a 'rights watch' to monitor any changes in legislation, as well as changes in the rights status of the material in the preservation repository's care (e.g. through the expiration of copyright).

Rights metadata in the DIP

The user of an archive will not need access to preservation-related rights information. However, it is important that they are given comprehensive access-related rights information, especially about the copyright status of an item. This should be recorded as part of the DIP for each digital object in an archive.

The DIP should include information drawn from the METSRights record relating to permissions and constraints on access and use; the contact details of the rights holder or their representative should be omitted from this public record because including these without first obtaining permission would contravene the Data Protection Act. A statement should be added in the <ConstraintDescription> field to indicate that users should contact the digital repository if they wish to seek permission from the rights holder for re-use of any copyrighted digital material.

Rights information could also be included in the Dublin Core record for a digital object for OAI-PMH harvesters. An ODRL/DCMI Profile Working Group has been set up, which plans to develop a way of making combined use of the rights-related DCMI metadata terms and the ODRL rights expression language. This will enable richer rights management information to be captured along with DCMI descriptive metadata.

Rights metadata in the EAD catalogue

The EAD catalogue provides the final, comprehensive layer of descriptive metadata, which enables the researcher to determine whether or not the content of an archive is relevant to their research. On identifying a relevant digital object, the researcher will be able to move from the catalogue entry to the DIP via a link. The DIP, rather than the EAD catalogue, should contain the principal rights metadata record for the researcher because:

The primary rights-related information in the EAD catalogue should occur at collection level; in a hybrid archive this will draw together relevant rights information about both the paper-based and digital components of an archive. An outline of suggested content for the <accessrestrict> and <userestrict> elements at collection level is given in Chapter 06 Arranging and cataloguing digital and hybrid archives. At lower levels, a brief <accessrestrict> statement can be used to indicate if component items are closed or subject to access restrictions.

Rights metadata and controlling access

Most of the material in a personal digital archive will remain in copyright for many years and will only be made accessible in a managed searchroom environment. Sometimes, however, it may be possible to obtain a licence from a rights holder authorising the repository to make material available to researchers remotely, with certain conditions attached, e.g. access might be limited to authorised remote users or to a restricted number of computers; or there may be limits on the number of each item which can be printed. In this case, digital archivists may choose to employ rights expression languages which could be embedded in software to control usage of material. Coyle, for instance, suggests that the Federated Digital Rights Management (FDRM) architecture could in future be developed into a system for enforcing access rights based on licence statements like those recorded in METSRights.

However fully a repository documents the IPRs associated with the digital objects in its collections, and however effective the repository's rights management policy, Paradigm also recommends publishing an institutional 'take down' policy, whereby material can be removed from public access at the request of a rights holder. This is crucial in the context of personal archives, where large numbers of third party rights holders are represented in collections and explicit permission cannot be sought from each to copy or make available the documents to which they hold the rights. Having a comprehensive rights management policy does at least demonstrate that a repository is acting responsibly, obtaining permission where possible and complying with the relevant legislation where this is not possible.