Versioning the Dublin Core across multiple languages and over time

Abstract

The Dublin Core Metadata Initiative is currently designing and implementing a registry for managing its ofslcial element and qualifier definitions over time and in multiple languages. Three aspects of this registry must be versioned: the schema descriptive of elements and qualifiers itself, which may be accessed by software or Web agents; sets of elements and qualifiers as wholes, of which some software vendors require stable snapshots as targets for development; and individual elements and qualifiers, which in their life cycles may change in meaning or approval status. Each such event must be uniquely identifiable in a persistent way, and translations must point to the precise sources on which they were based. Versioning a metadata language such as Dublin Core over time is a balancing act between an error-tolerant fuzziness and a desire for archival precision. This is consistent with the philosophy of a metadata pidgin that aims at universal interoperability while recognizing the limits of the goal itseg Registries as dictionaries of metadataMetadata is a key aspect in the design of digital libraries. Very broadly, metadata is "(structured) data about data." In specific application areas, this may include data for describing resources, controlling access to intellectual property, specifying terms and conditions of use for Internet commerce, rating the quality or suitability of contents for particular audiences, and managing access to archived contents over the long term. Maintaining digital library services that are interoperable, reliable, and future-proof will require an infrastructure for documenting metadata terms as they are defined and change over time.Metadata terms are defined and published in "schemas," and schemas may fulfill a wide variety of functions. Schemas defined by standardization committees and published for general use are called "namespaces," and namespaces translated into languages other than the (usually English) original are "translated namespaces." At the other extreme, schemas that use terms from existing namespaces and combine or customize those terms for a particular application are called "application profiles." Like the Web itself, Web-based metadata schemas are a relatively new phenomenon, so this simple typology may complexify over time.In this paper, software systems that provide human-and machine-readable access to these various types of schema are called "metadata registries." Structured documentation on schemas will be important for software developers, content providers, and end users, and access to machinereadable schemas will be crucial for the smooth operation of automated Web applications and software agents. A joint NSF and European Union Working Group on Metadata recently predicted the emergence of an ecology of registries -"An ecology of registries will emerge that reflects a diverGty of organizational motives, market forces, and user requirements. A distributed model for registries will reflect this diversity by supporting cooperation at multiple levelsglobal, regiona...

A distributed archival network for process-oriented autonomic long-term digital preservation

“…The goals of the Dublin Core team are: simplicity, interoperability, applicability, extensibility, adaption of standards and common semantics. The Dublin Core is great for making basic statements about resources and it is a small language [32]. The Dublin Core exists of two categories: elements and qualifiers [33].…”

Dublin Core: Process and Principles

“…Every DCMI term has its approval status and human readable label and descriptions, which could change over time. Every term could have translations which may be appended and modified over time [11]. Maintenance of local or domain specific schemas is also an important and challenging issue because the community maintaining the registry has to maintain consistency with other registries such as the central DCMI registry.…”

DCMI metadata schema registry for sharing authoritative information about metadata schemas

“…A metadata schema registry is widely recognized as an important tool not only to share information about metadata vocabularies but also to enhance reusability of metadata vocabularies. A registry also plays important roles in facilitating semantic metadata interoperability among communities speaking different languages and over time [1]. Achieving metadata interoperability is fundamental to making information resources shareable and discoverable.…”