A beta version of the Core Vocabularies / Data Model tool has been released for trial. The address of the beta is: https://tietomallit.beta.yti.cloud.dvv.fi/

  • The landing page displays published data models, which are either in Valid or Proposal status. Data models in Draft status are only visible to people in your organisation. 
  • The Beta tool uses the same access rights information as the current tool. 

The new tool has the same functionalities as the old one plus a number of new ones. The changes and new features are described in more detail below.

User interface and accessibility

  • The interface of the tool has been redesigned for accessibility.
    • The interface has been implemented using DVV's Design System components, which themselves are accessible.
  • The data model has its own search function to retrieve the resources in the data model, i.e. classes, attributes and associations; formerly the data model resources could only be browsed, but not searched.

Visualisation (graph diagram)

  • Core data models (formerly data components) and application profiles are now visually different:
    • the graph of the core data (visualisation) is purple and
    • the graph for the application profile is blue.
  • Associations or attributes can also be added to a class via the visualisation.
  • The graph now also shows the links to the superclass, i.e. the superclass-subclass relationship, so there is a new link type in addition to the former association relationship.
  • In the visualisation, you can toggle to display either technical names or human-readable names. Also attributes and associations can be hidden in the visualisation.
  • A class borrowed from another data model (i.e. a class previously added using the 'Use class' function) is now only displayed as a link in the visualisation. In practice, they are references to a class in another data model.

Principles of modelling

  • The distinction between core vocabularies and application profiles has been clarified and the role of core vocabularies strengthened. The content of the application profiles should be based on the core data, whereas previously it was possible to freely invent classes to be used in the application profile only.
  • Previously, the tool did not support the use of data from another application profile. Now the resources in an application profile can also be based on resources in another application profile. This is done by copying their references to core vocabularies. So you can add to your own data model the same kind of reference to the core vocabulary as there are in the resource to be copied.
  • The tool better supports the use of external resources: the data models can be based on core data vocabularies that are not in the Interoperability Platform. This is possible if they are linkable (resolvable) ontologies, such as the Europass Learning Model Ontology (ELM). Established, resolvable core vocabularies and ontologies can be added to the tool as so-called built-in namespaces.
  • Earlier, it was mandatory to link all vocabularies, terminologies and code sets (enumerations) to your data model before starting modelling. Now you can add them manually before starting modelling or the tool can add them automatically, as you add resources to your data model.

Identifiers

  • Addresses, i.e. identifiers will be IRI addresses instead of former URI addresses. 
    • IRI addresses are an extension of URI addresses. URI addresses can only use ASCII characters, while IRI supports also non-English characters (i.e. UTF-8 encoding).
    • The address will be like https://iri.suomi.fi/model/isa2core/, whereas it is now http://uri.suomi.fi/datamodel/ns/isa2core#
    • For the time being, the addresses in the tool are generated according to the principles of URI addresses.
  • After migration of data from the old tool to the new Data Model tool, the old URI addresses of the old tool will be redirected to the new IRI addresses.

Versioning

  • The versioning of the data model has been renewed.
  • Previously, the steps in the modelling process were expressed only with status codes, but now the published data models are given not only status codes but also a version number.
  • The entire data model has the same status - it is no longer possible for a data model status to be Valid, but have resources within it that are, for example, in the status Retired.
  • There is a working version of the data model having the status Draft. It does not have a version number. This working version is only visible to your organisation.
  • The public data models are published as snapshots from the working version. The content of these published data models cannot be changed. So other parties can safely link to them, as this data will not be removed and cannot be changed.
  • The IRI address of the data model is the same for all versions. When linking, it can be indicated whether the link is to the data model in general or to a specific version of it.

Key features not been implemented (yet)

  • Some technical formats (serialisations) are still missing. Only JSON-LD, RDF and Turtle are now ready. Possibly we will implement Open API, if there is a demand for it. 
  • The notification service is not yet available, i.e. it is not yet possible to subscribe email notifications about new versions of the data model.
  • The Swedish language interface is not yet available.



  • No labels