Skip to main content

CMDI 1.2 for metadata modellers

CMDI 1.2

Information for metadata modellers

Metadata modelling is the task that logically precedes the metadata creation task. The parts of the CLARIN metadata infrastructure that are involved are the CMDI Component Registry, CLARIN Concept Registry and CLARIN Vocabulary Service (CLAVAS).

What is CMDI 1.2?

See the page on CMDI 1.2 for general information about 1.2.

How does CMDI 1.2 affect me?

As a metadata modeller, you will notice a number of small but significant changes in the CMDI Component Registry starting with version 2.1, the first version to support CMDI 1.2. All component and profile specifications have been converted to CMDI 1.2, which becomes apparent when you view the XML (for example using the 'xml' tab in the component browser part of the user interface) of any item in the registry. The drop down menu for individual items and the 'Show info' dialogue now show download and link options for both CMDI 1.1 and CMDI 1.2. In the editor, a number of new options have become available, all of which only affect the CMDI 1.2 versions of the derived profile schemata:

  • Attributes have a checkbox to make a specific attribute required in the instance
  • There now is a documentation field for components and attributes in addition to the one for elements (and multilingual documentation will become available in Component Registry 2.2)
  • The option to link external vocabularies to elements, either as 'open' or 'closed' vocabulary (will become available in Component Registry 2.2)
  • Options for specifying cues for tools and means to element derive values from other elements (will become available in Component Registry 2.2 or later)

With CMDI 1.2, a new infrastructure component has entered the realm of component metadata: the CLARIN Vocabulary Service. CLAVAS is a platform for SKOS (Simple Knowledge Organization System) vocabularies and has been established to provide centrally maintained (i.e. not open for arbitrary editing) controlled vocabularies that can be linked to from element (field) declarations in metadata definitions. In CMDI 1.2, a vocabulary can be applied as a closed vocabulary, i.e. defining the finite set of allowed values for that element - or alternatively as an open vocabulary, in which case it serves as a hint to editors and other CMDI processing software to suggest or prefer values for the linked vocabulary. CLAVAS vocabularies are not (per se) static, but closed vocabularies will be frozen and integrated into component or profile once linked to allow for schema validation. In the future CLAVAS may be extended with a versioning mechanism through which specific static revision of a vocabulary can be used as open vocabularies but currently this is not an option. To experiment with versioning and ways to maintain vocabularies the CLARIN Centre Committee has started a pilot on managing vocabularies in CLAVAS. If you want to get in involved please send a mail to cmdi [at] clarin.eu (cmdi[at]clarin[dot]eu). Also synchronized with the Component Registry 2.2 release CLAVAS will be updated to 2.0 and its vocabularies better aligned with the recommended components in the Component Registry.

Another feature related to CMDI 1.2 that will become available (but will apply to registry items in general, including 'legacy' CMDI 1.1 components and profiles) is lifecycle management. It allows metadata modellers to mark components and profiles as deprecated, and optionally appoint successors upon deprecation. It will also be possible to make development version of items publicly visible, i.e. "pre-publish" a component or profile while retaining the option to make modifications. These features are expected to be implemented in the upcoming Component Registry version 2.2.

Since enabling the creation of metadata is the ultimate goal of the metadata modelling task, please also read the section for metadata creators above.

Should I take action immediately?

As a metadata modeller, you do not have to take any specific action and can keep using the Component Registry's user interface to create components and profiles. When creating new components or profiles it's always safe to use the new CMDI 1.2 features (see above), that is, it will always be possible for metadata creators to use it with CMDI 1.1 metadata. However, keep in mind that in those cases some features will not be reflected in the schema or instance.

How can I switch to CMDI 1.2?

All components and profiles in the Component Registry have been converted and new items will be stored as CMDI 1.2 specifications as well. Therefore no effort is required from metadata modellers to switch to CMDI 1.2 other than making use of the newly added features in new or derived components and profiles. For example, one can use the "edit as new" functionality to make a new version of an existing component and link open vocabularies to one or more fields (have a look at the vocabularies currently available in CLAVAS) or improve its usability by adding documentation to components and attributes. The old version of such a component can then be deprecated and the new version appointed its successor.