DOIs (Digital Object Identifiers) enable unique, persistent identification for objects as an important part of the open research infrastructure that helps to make research and scholarly communication more FAIR (Findable, Accessible, Interoperable, and Reusable). If your organization is thinking about using DataCite DOIs, or if you’re just getting started, this blog provides an overview of best practices and important considerations to keep in mind as a responsible member of the persistent identifier and open research ecosystem. 

DataCite DOI Registration Policy

As a starting point, all organizations creating DataCite DOIs should be aware of DataCite’s DOI registration policy:

  • Only assign DOIs to content that your organization has responsibility for (hosted/managed/created at your organization).
  • Only assign DOIs to unique content - do not assign DOIs to content already published somewhere else with a DOI.
  • Please check copyright before assigning a DOI.
  • DOIs must resolve to a publicly available landing page. The underlying content does not need to be publicly available but the metadata must be open.
  • Update DOI metadata if it changes over time.

You will need to make sure your organization can adhere to this policy prior to creating DataCite DOIs.

DataCite DOI Metadata

DOIs must resolve to a public landing page that contains information about the object, with instructions on how to access the object itself. Every DOI must have a minimum amount of mandatory metadata describing the object that the DOI is assigned to, including: Creator(s), Title, Publisher, Publication year, and Resource type. DataCite DOIs can be assigned to many different types of resources, ranging from datasets to audio-visual materials, images, collections, gray literature, software, publications, data management plans, and more. See the latest version of the DataCite metadata schema for a complete list of resource types and metadata fields that can be associated with a DOI. The more metadata included in a DOI, the more useful the DOI will be towards making research and scholarly materials more FAIR: Findable, Accessible, Interoperable, and Reusable in accordance with the FAIR Principles

Other Persistent Identifiers (PIDs) should be included in DOI metadata in order to persistently connect objects with their creator(s), affiliated organizations, related objects, and funding for the object’s creation, if applicable. These other identifiers should be included in DOI metadata as follows:

Once a DOI is registered, the metadata behind the DOI can be changed as needed, especially the URL or location of the landing page that the DOI needs to resolve to. It is especially important to make sure the URL associated with the DOI remains accurate, as well as generally keeping DOI metadata up to date. If multiple versions of a resource exist, each version can have its own DOI, with all of the versions’ DOIs linked together via the Related Identifiers metadata field. Several platforms harvest DataCite metadata directly for use in search or analytics features, including Clarivate’s Data Citation Index, Mendeley Data, Dimensions, Ex Libris Primo, and Proquest Summon. DataCite metadata is also harvested into Scholix , which is a meta-service for linking publications to datasets. Metadata for DOIs that are labeled as datasets is also indexed by Google Dataset search and OpenAire. All of these discovery services work best when descriptions, subject terms, and related identifiers are included in the DOI metadata.

Digital Preservation

Just as DOIs should enable a persistent way to find an object, the objects for which DOIs are created should also be persistent. While it is possible to assign a DOI to a physical object, most DOIs are assigned to digital objects. Either way, preservation of these objects is a key part of maintaining the integrity of DOIs. Any digital object that a DOI is assigned to should be part of a digital preservation program to ensure that the object itself continues to be accessible over time.

Digital preservation programs should ideally allow for three digital copies on different media formats, stored in different geographical areas, with regular bit-checks to check for bit integrity. There are several digital preservation solutions available, but some of the open source options include: DuraCloud, Archivematica, and LOCKSS. For resources to learn more about digital preservation best practices, please see:

Get Involved

The Lyrasis DataCite US Community provides a consortium option for organizations that are new to DataCite; please use our contact form to express your interest or ask questions.