skos:ConceptScheme

As a basis for classification might be used, for example, [[TERMDAT]], [[PLZO_CH]], or a growing terminology gathered by a portal itself. Terminologies determined by the eCH Standards might also be appropriate (for example eCH-0049, eCH-0070 and eCH-0145 for describing the services or tasks of the public administration.)

For exchange with other European portals, the upper level of the multilingual EU thesaurus [[EUROVOC]] is supported.

The element dct:title can occur more than once, in order to express the title in several languages.

Multilingual titles don't mean that the data of the dataset is available in those languages. But if the dataset has a physical representation in a language it MUST also come with a title in that language. This way a title can be given to the physical representation even if the physical representation does not come with a title itself ().

This property names the organization responsible for the publication of the dataset. If there are more organizations responsible, the property can be repeated.

If a [[TERMDAT]] entry exists for the publisher, this identification will be entered additionally. In case the publisher only provides the TERMDAT identification, the portal software determines the corresponding publisher name via the TERMDAT database. If only the publisher name is provided the portal software adds the TERMDAT identification from the database, in case it can be found. Otherwise, the identification is left empty. If the publisher provides both the TERMDAT identification and the name, the portal software accepts both values without changing them.

Internally the property will be represented in the following way:

It is necessary to uniquely mark each Dataset in order to be able to link it to similar datasets. The identifier of a dataset should therefore have the following structure:

ID_Datenset@ID_Datapublisher

The data provider selects a data provider identifier which, based on human discretion, is nationally unique.

Datasets will be linked via the element "rdfs:seeAlso" by specifying in each case the corresponding element dct:identifier>.

Using rdfs:seeAlso for linking to other datasets makes it necessary that a data portal implementing this automatically takes care of consistency if datasets get changed or removed:

• If the identifier of a dataset is changed: all seeAlso properties of other datasets on the portal should be changed accordingly
• If a dataset is deleted: its identifier should be removed of the seeAlso of other datasets.

It is good practice to mark the language of the keywords with the [[ISO 639-1]] language code such as "geodata"@en.

This property refers to a geographic region that is covered by the Dataset. Possible values for Swiss datasets might be the official canton abbreviations or community names as they are published by swisstopo ([[CADASTRE-PLZ-CH]]).

Another option is to provide point coordinates or a rectangular bounding box. It is not possible with dct:Location to provide a polygon as geographical coverage, therefore dct:coverage should be used for this purpose ()

The property should only be used for spatial coverage that has the shape of a polygon. (for example name="dc.coverage.polygon" scheme = "WGS84.DD" content = "9.547 46.737 9.550 46.739 9.556 46.740 9.554 46.743 9.548 46.742 9.547 46.737"; name= "dc.coverage.placeName content="Heidsee"). In this case there must a property dct:spatial added as well in order to stay conformant with DCAT-AP. The portal software will do that automatically.

With this property, a dataset can link to a related dataset.

Additional material such as an article, example or documentation helpful in the context of the dataset can also be linked in this way.

The element is also used to link to legal foundations that apply to the publication or usage of the dataset.

The element rdfs:seeAlso enables data publishers to add a reference, in the metadata of a dataset, to another dataset. This allows users to discover datasets related to one they have already found.

The element rdfs:seeAlso is added to the metadata of a dataset as necessary. In this way, related datasets (e.g. population data from 2013 and population data from 2014) can be simply connected with one another.

Internally, the dct:identifier property of the related dataset will be used.

If the dataset is only accessible via a landing page, then the value entered in dcat:landingPage should be the same as the value in dcat:accessURL.

Since the field is mandatory in the DCAT-AP it has to be set even for downloadable distributions. In that case both dcat:accessURL and dcat:downloadURL should contain the same value.

There might be more than one way to access a distribution (HTTP, WebDAV, FTP) there might be several access urls for a distribution.

To specify the rights statement for a distribution on a Swiss open data portal, the following information is necessary:

• Attribution: required/not required
• Non-commercial use: allowed/not allowed
• Commercial use: allowed/not allowed/only with approval

The element dct:rights should be formed with the vocabulary [[ODI-VOCAB]]:

The individual conditions are formulated as follows:

• ReferenceRequired / ReferenceNotRequired
• NonCommercialAllowed / NonCommercialNotAllowed
• CommercialAllowed / CommercialNotAllowed / CommercialWithApprovalOnly

The fully-specified rights statement is then assembled from the three relevant conditions:

In case attribution is required, the text to be used must additionally be included:

The data owner must also be able to list the legal basis for the collection and publication of the data. For this, the element dct:relation may be included in the element dct:rights, with the following form.

The title MUST be given if the distribution contains only part of the data offered by the dataset. .

The title can be given in several languages. In multilingual data portals, the title in the language selected by a user will usually be shown as title for the distribution.

The description MUST be provided if the distribution contains only part of the data offered by the dataset. .

The description can be given in several languages. In multilingual data portals, the description in the language selected by a user will usually be shown as description for the distribution.

The value of the media type MUST be a IANA MIME type, see [[IANA-MEDIA-TYPES]].

The property MUST be set if the distribution is language-dependent, or if it is given in some of the languages German, French, Italian and English but not in all four languages.

If there is a distribution with a language property set, the dataset MUST also provide a title in that language ()

In the element dct:license, one license can be specified. It is recommended to use a current, international standard license, e.g. the CC-0 Creative Commons Public Domain Dedication.

This differs from the element dct:rights, where any information necessary to specify the usage rights can be added.

Because the distribution might be accessible via various methods (HTTP, WebDAV, FTP), there MAY be more than one download url.

In order to ensure the conformance with DCAT-AP each download url must be doubled as access url.

This is the key element for supporting spatial or temporal range searching on document-like objects that are spatially referenced or time-referenced.

The content of various distributions in a dataset can differ from one another (e.g. population data for different years). For this reason, the specific data in the distributions may differ from the general information in the metadata of the dataset, particularly in terms of time and location.

In this case, the element dct:coverage, which can contain both time and location information, is permitted in DCAT-AP CH-compliant metadata.