The DOI Foundation: Products and Identifier Technologies

A DOI is a digital identifier of an object — physical, digital or abstract — designed to solve the problem of keeping track of things, whether matter, material, content or activities. A DOI name is a unique number made up of a prefix and a suffix separated by a forward slash, such as 10.1000/182, and prefixes always start with "10.". DOIs are designed for use by machines as well as humans and act as persistent identifiers, providing a long-lasting reference to an object or to information about it. They are also actionable: prefacing a DOI with doi.org/ turns it into a link that resolves to current information about the object.

• Prefix and suffix separated by a forward slash
• Prefixes begin with 10.
• Applies to physical, digital or abstract objects
• Persistent and actionable
• Free to resolve

The DOI system is a generic framework that provides an actionable, interoperable and persistent link for any entity, described as a "Digital Identifier of an Object" rather than an identifier of a digital object. It is actionable through identifier syntax and a network resolution mechanism, persistent through handle infrastructure combined with obligations placed on Registration Agencies, and interoperable through a data model supporting semantic interoperability and grouping. The system complies with ISO 26324, Digital Object Identifier System, published 1 May 2012, and the DOI Foundation is the ISO registration authority for it. Extensible by design, the system began with documents and media such as articles and data sets and now also covers parties and licences.

• International standard ISO 26324 (1 May 2012)
• Foundation launched 1998; first applications launched 2000
• Uses the Handle System as its resolution mechanism
• Extensible to any sector, not publishing-only
• shortDOI service for shortened links to DOI names

DOI names structured for the DOI system Proxy Server are resolved through https://doi.org, the preferred address; the earlier dx.doi.org syntax continues to be supported. The proxy server is a web server that communicates with the Handle System: for an HTTP request combining the proxy domain with a DOI name, it queries the Handle System, takes the URL in the handle record and issues an HTTP redirect to the user's browser. Some DOI names carry data in addition to a single default URL, known as multiple resolution, intended for applications that can use enhanced metadata or related documents. The proxy service is accessible over IPv6, supports DNSSEC, responds to HTTPS and HTTP, and offers a REST API alongside documented query parameters and URI encoding rules.

• Preferred proxy address https://doi.org
• Support for handle value type 10320/loc multiple redirection
• IPv6 and DNSSEC support
• Proxy Server REST API
• "DOI Name Not Found" error page for unknown names

The DOI Data Dictionary is a published reference of terms used within the DOI system, arranged alphabetically with definitions for entries such as AbbreviatedName and AbbreviatedTitle. The DOI system's semantic interoperability model is based on work derived from the indecs Framework, an analysis of metadata requirements for e-commerce of intellectual property that addressed how one system can understand terms used by another. indecs supports interoperability across media, functions, levels of metadata, and semantic and linguistic barriers, and rests on principles including unique identification, functional granularity, designated authority and appropriate access. DOI metadata schemes are among the activities that have used indecs.

• DOI Data Dictionary, version dated 2015-03-23
• Alphabetical listing of metadata terms and definitions
• Semantic interoperability model based on indecs
• Related work includes the Vocabulary Mapping Framework and ONIX

OpenURL is a NISO standard syntax for transporting metadata and identifiers about one or more referents within URLs. It provides a syntax for encoding metadata but is not itself a source of metadata, and it is restricted to the world of URLs, whereas the DOI system has wider application. The DOI Foundation describes the two as complementary: the DOI system resolves identifiers to global services, while OpenURL contextualises requests to those services for local requirements. Used together, they allow a richer user experience that combines global and local needs, with the DOI system supplying identifiers and authoritative metadata and a single network point for creating and redirecting OpenURLs.

• OpenURL is a NISO standard syntax
• Syntax for encoding metadata, not a metadata source
• Addresses the appropriate copy problem
• Complementary to DOI resolution to global services

A PURL is a Persistent Uniform Resource Locator: functionally a URL, but pointing to an intermediate resolution service that associates the PURL with the actual URL and returns it as a standard HTTP redirect. The OCLC PURL Service was shaped by OCLC's participation in IETF Uniform Resource Identifier working groups, and PURLs were developed as an approach to the problem of unstable URLs. The DOI Foundation notes that PURLs provide one level of indirection contained within a single server permanently attached to a specific domain name, and that PURL servers do not know about each other. The redirection is functionally equivalent to the way the DOI system uses its proxy server, but the DOI system adds multiple resolution and data types, a metadata approach, enforced common policy and centrally managed redirection.

• PURL resolution returns the target URL as an HTTP redirect
• All PURLs are HTTP-based
• Single level of indirection within one server
• DOI adds multiple resolution, metadata model and common policy

ISBN-A, "the actionable ISBN", is a service powered by DOI in which an existing ISBN is expressed within the DOI system as a DOI name derived from that ISBN. It can only be registered for a 13-digit ISBN; legacy 10-digit ISBNs must first be converted, usually by adding the 978 GS1 prefix and recalculating the check digit. An ISBN-A identifies the same referent as the ISBN it is derived from, but ISBN-As do not exist automatically for every ISBN and are assigned only by DOI Registration Agencies authorised to do so. The purpose is to make an existing ISBN useful on digital networks, for example by linking a publication's identity to a quality-controlled destination, supporting variable licensing or tracking digital product sales.

• Derived from a 13-digit ISBN only
• Assigned by authorised DOI Registration Agencies
• Identifies the same referent as the source ISBN
• An ISBN alone cannot be resolved in the DOI system

DOI names are obtained through a service offered by a DOI Registration Agency, which collects metadata, assigns DOI names and offers further services such as reference linking or metadata lookup. Membership of the DOI Foundation is not required in order to work with a Registration Agency. Registration costs depend on the services purchased, and each agency is free to offer its own business model; existing DOI names remain free to resolve. Organisations that can represent a defined community of interest may apply to the DOI Foundation to become a Registration Agency.

• DOIs are assigned via DOI Registration Agencies
• Membership of the Foundation not required to use an RA
• Resolution of existing DOI names is free
• Contact address info@doi.org