1 Purpose

This statement describes the core policies underpinning the ARDC DOI service. Most of these policies arise as a consequence of the design and architecture of the global DOI and DataCite infrastructure which supports this service.

2 Background

DOI Service is an ARDC service that provides cost free creation (called minting) and support for user updating of Digital Object Identifiers (DOIs) for the Australian research datasets, collections or grey literature. See more information on the DOI Service. General information on data citation is also available at Data Citation - ARDC Guide.

3 Policies

3.1 Service functionality

The ARDC DOI Service will allow research organisations to assign and manage Digital Object Identifiers (DOIs) for research datasets, collections or grey literature. The DOI Service functionality is built upon external DOI services offered through our membership with DataCite. This adoption of external services means that the functionality available through the ARDC DOI Service is somewhat driven by the functionality exposed by DataCite.

3.2 Service scope

The ARDC DOI Service is designed for use by Australian research organisations or data centres relevant to Australian research, such as those run by Government agencies and universities.

In-scope:

• Research datasets and collections, associated workflows, software and models
• Grey literature such as theses, reports, unpublished conference papers, newsletters, creative works, preprint journal articles, technical standards and specifications for which the institutional repository is the primary publication point

Out of scope: 

• published peer reviewed journal articles, ephemera, teaching and learning materials and book chapters

3.3 Service access

Clients are required to register with ARDC before access to the ARDC DOI administrative services are granted. Once registered, clients can programmatically call the DOI machine-to-machine (m2m) service to mint and manage DOIs. An interface is also available to registered clients for manually minting DOIs.

3.4 Data openness

The DOI Service is intended to identify datasets, collections or grey literature being made available freely to the national and international research community as part of the Australian Research Data Commons. Datasets, collections or grey literature to which DOIs will be assigned should:

• Be accessible. Note this can include both open access (example) and mediated access (example).
• Be part of the scholarly record.
• Be persistently available.
• Have the metadata required by the DOI Service.

ARDC assumes no responsibility for the selection of datasets, collections or grey literature which are to be assigned DOIs. This remains the responsibility of the institution which has ownership of the dataset, collection or grey literature.

3.5 Scholarly record

The datasets, collections or grey literature assigned DOI's by the ARDC DOI Service should be citable contributions to the scholarly record, analogous to a journal article.

3.6 Service style

The ARDC DOI Service is a set of m2m web services intended for use by application developers and research data management system users.

3.7 DOI structure

Production DOI's minted by the ARDC DOI Service will take the form 10.4225/suffix, 10.4226/suffix or 10.4227/suffix where suffix is a random string of digits. Test DOI's minted by the ARDC DOI service will take the form 10.5072/suffix where suffix is also a random string of digits.The DOI for each institution will have a standard identifier as part of their DOI numbers which will allow them to more easily track reuse of the citations.

For example:

   Figure 1: Anatomy of a DOI

This is an example from Geoscience Australia:

3.8 DOI metadata

DataCite has published an XML schema that defines the valid format for any metadata submissions to the dataset DOI registration system. As the ARDC DOI Service integrates with the DOI system via DataCite, all metadata that is passed as a parameter to one of the DOI web services must adhere to this schema. For more information on the schema please refer to the DOI Service Technical Documentation m2m service. Once a DOI has been minted, the ARDC DOI Service will give public access to the metadata associated with the DOI via a publicly accessible service point. The metadata will also be publicly visible through the DataCite Metadata Search.

3.9 Management of DOIs

ARDC assumes no responsibility for the management of DOIs minted using the ARDC DOI service. It is the responsibility of the client who minted a DOI to ensure it is kept resolvable, and the metadata associated with a DOI is current.

3.10 Other data items

The datasets, collections or grey literature assigned DOIs by the ARDC DOI Service can take the form of most digital data outputs, with the exception of concepts or people and organisations. These may be better identified using parallel identifier standards for these entities, e.g. the ARDC Handle service.

3.11 Labels

3.11.1 Labels and meaningfulness

Labels for persistent identifiers can allow a user to infer things about the object being identified from the label itself. This can make the identifier easier to remember, easier to enter without error and easier to communicate to others. However, meaningful labels are usually based on attributes of the objects identified that are less likely to persist than the object itself. Meaningful labels then become misleading or possibly broken.The ARDC DOI Service does not allow clients to allocate their own identifier labels. The service generates and assigns non-meaningful numeric identifiers as suffixes to DOIs minted through the DOI Service.

3.11.2 URL safety

Persistent identifiers are often used as part of URLs. They should not contain characters which need encoding to be embedded safely in URLs, such as '&' or space: such conversion can confuse users as to whether the encoded or the unencoded label is the 'real' label. For example, 'a&b', when URL-encoded, becomes 'a%26b'.The ARDC DOI Service always generates URL-safe identifiers.

3.11.3 Variant forms of labels

Persistent identifier labels with multiple possible variant forms should be avoided, as users (or systems) risk assuming that the variants are distinct after all. Case sensitivity should be avoided, as should visually confusable characters (1 I l, 0 O), as humans risk failing to distinguish them. The ARDC DOI Service does not allow clients to allocate their own identifier labels. The ARDC DOI Service generates labels which cannot be confused within ASCII, as it uses numeric labels exclusively.

3.11.4 Punctuation within labels

Labels will likely be delimited by punctuation, both when cited in running text, and when embedded within URLs or other identifiers. Consequently, punctuation should be avoided in labels. The ARDC DOI Service does not use punctuation within labels.

3.11.5 Label length and format

As labels will sometimes need to be written down or manually entered into a system, they need to be short enough to write down or to type. The DOI Service uses short numeric labels.

3.12 Service availability

ARDC will endeavour to ensure that the services underpinning the ARDC DOI Service are highly available. However, occasional maintenance outages are inevitable. Users of the DOI Service should ensure that software being integrated with this product is designed to cope with service unavailability.