opendata-swiss / dcat_ap_ch Goto Github PK
View Code? Open in Web Editor NEWExamples for geocat and DCAT data-catalogs are given here
Examples for geocat and DCAT data-catalogs are given here
Property: all properties of DCAT-AP that are not included in DCAT-AP CH
Class: all classes of DCAT-AP that are not included in DCAT-AP CH
Conformance Problem: Data receivers(e.g. opendata.swiss) who have DCAT-AP-CH as a reference should still be able to pass through well formed DCAT-AP catalogues, even if some properties or classes are not directly defined in DCAT-AP CH. In particular, data exported toward data.europe.eu should match the import of a data provider who chooses to use further DCAT-AP classes or properties.
This issue has been discussed and clarified here: SEMICeu/DCAT-AP#198
Proposal:
Possible consequences on opendata.swiss:
Par. 2.3 Multilingualism
First bullet point
Property: dct:publisher
Class: dcat:Catalog and dcat:Dataset
Conformance Problem: DCAT-AP 2.0.0 recommends a vocabulary for the publisher
Details: "The Corporate bodies NAL must be used for European institutions and a small set of international organisations. In case of other types of organisations, national, regional or local vocabularies should be used"
Proposal: A vocabulary has been proposed: https://www.dcat-ap.ch/vocabulary/publishers/20210623.html: the entries for it have been extracted via a tool from the names entered in dct:publisher on https://ckan.opendata.swiss
Property:dcat:theme
Class: dcat:Dataset
Conformance Problem:
Details: The EU vocabulary to use is this: http://publications.europa.eu/resource/authority/data-theme
Proposal:
Property: dcat:landingPage
Class: dcat:Dataset
Implementation Problem:The implementation on opendata.swiss does not conform this DCAT since dcat:landingPage is implemented as an rdfs:Literal instead of a foaf:Document
Details: .Implemented:dcat:landingPage: "https://swisstopo/index"
, DCAT-AP CH conformant: dcat:landingPage:<https://swisstopo/index>
Proposal:Change the implementation on opendata.swiss so that it conforms to DCAT-AP-CH
Properties: dct:modified
Classes: Dataset, Distribution
Ist: The actual guidance about when should a new value for dct:modified should be set is insufficient. https://dcat-ap.ch/releases/2.0/dcat-ap-ch.html#dataset-modification-date
Soll: The eCH-Working Group defines clearer criteria which defines which actions should be counted as a change to the dataset or to the resource.
Property: dcat:format
Class: dcat:Distribution
Conformance Problem: DCAT-AP differs in the requirement level of this property: it is recommended wheras dcat:mediaType is optional
Proposal: Adapt the usage note and change the requirement level to recommend
Property: rdfs:seeAlso
Class: dcat:Dataset
Conformance Problem: rdfs:seeAlso
is a custom Swiss property to express relationships between datasets. DCAT already includes properties for this purpose, however. DCAT gives the choice between dct:relation
and dcat:qualifiedRelation
. In order to conform to DCAT, one of those properties MUST be used instead.
Proposal: Abondon the property and use DCAT vocabulary dct:relation
or dcat:Relationship
instead.
Property: dct:publisher
Class: dcat:Dataset
Conformance Problem: Currently DCAT-AP-CH allows more than one publisher per dataset, but DCAT-AP 2.0.0 allows only one publisher
Proposal: change the cardinality so that it conforms to DCAT-AP
In the German DCAT-AP 2.0 standard the attribute dct:references
is used to link to their "Musterdatenkatalog", a catalog of reference datasets. The goal is to provide a way to find similar datasets across different portals.
E.g. "Baumkataster" has the following reference dataset: https://musterdatenkatalog.de/def/musterdatensatz/gruenflaechen/baumbestandBaumkataster
If this property is adapted, we need to make a note about its implementation. We could either create a new Swiss-specific "Musterdatenkatalog" or use https://musterdatenkatalog.de of Bertelsmann.
dct:reference
in the German specification: https://www.dcat-ap.de/def/dcatde/2.0/spec/#datensatz-referenziertProperty: dct:mediaType
Class: dcat:Distribution
Conformance Problem: The media type must use IANA media types as a controlled vocabulary. Media types should be provided as URIs.
Details:
Proposal:
Property:dcat:byteSize
Class:dcat:Distribution
Implementation:as Intenger dcat:bytSize: 1234
Conformance to DCAT-AP-CH: dct:byteSize
is demanded as rdfs:Literal
typed as xsd:decimal
in both DCAT-AP CH and DCAT-AP. So the example should be implemented as dcat:bytSize: "1234"^xsd:decimal
Conformance to DCAT-AP: DCAT-AP-CH conforms to DCAT-AP in this regard
Proposal: Change the implementation on opendata.swiss
Property: dct:rights
Class: dcat:Distribution, dcat:Catalog
Conformance Problem: The property is described and implemented incorrectly. It is not meant to replace dct:license but to add an information on the legal framework and thus complement dct:license
Details: See here https://theodi.org/article/publishers-guide-to-the-open-data-rights-statement-vocabulary/ for how to use dct:rights.
Proposal: Refactor defintion of dct:rights and adapt to DCAT-AP by making dct:rights recommended and dct:license mandatory
Properties: dcat:downloadURL
Classes: Distribution
Ist: In DCAT-AP-CH v1 the defined cardinality is "1", whereas in DCAT-AP it's "n"
Soll: Adapt to DCAT-AP and define cardinality as "n" in DCAT-AP-CH v2. Check possible implementation issues on opendata.swiss
Report on the SEMIC Webinar on Identifiers (17.3.2022)
About: SEMIC invited the European national portals to a webinar about the future use of identifiers in DCAT-AP: "DCAT-AP webinar on identifiers"
There had several proposals. We could vote at the beginning for between 2 meaning of identifiers:
The second option was adopted since 90% of the participants voted for that option
After that several Proposals were presented:
Proposal 1: DCAT-AP recommends that all datasets have exactly one dct:identifier
-> We already have that dct:identifier as 1:n on the dataset class: so maybe we should change that to 1:1
Proposal 2: The identifier should come with its own metadata: since dct:identifier is a literal, it can't by itself connect to the original datasource. Therefore it is recommended to add a adms:identifier so that the dct: identifier can be set into context:
dct:identifier "dek-gs-3@kanton-thurgau";
adms:identifier [
skos:notation "dek-gs-3@kanton-thurgau" ;
dct:creator <https://opendata.swiss/organization/kanton-thurgau> ; ] ;
there can be more adms:identifier, but at least one of those should further describe dct:identifier
-> we don't have adms:identifier
on the dataset class yet, so we should consider to add this.
Proposal 3: The dct:identifier should be the IRI of the dataset:
<https://data.tg.ch/catalog/datasets/dek-av-23> a dcat:Dataset ;
dct:identifier "https://data.tg.ch/catalog/datasets/dek-av-23";
adms:idetifier [
skos:notation "https://data.tg.ch/catalog/datasets/dek-av-23";
dct:creator <https://opendata.swiss/organization/kanton-thurgau>
-> We should consider to add this to the usage note, but it seems difficult to implement this: currently many Swiss datasets don't have a URI and if the have one, it is not the dct:identifier
of the dataset.
There was a fourth proposal, but that one was not very well received by the community. The idea was that harvesters should also add an adms:identifier
with dct:creator
the dataportal that did the harvesting.
I have especially asked for guidance on how to implement Proposal 3 and someone else asked for that as well.
There will be a next Webinar on the topic. And I will mention it here, so that eCH Fachgruppen memeber, that are interested to attend will be able to ask SEMIC for an invitation.
Property: DCAT-AP properties that are not specified in DCAT-AP CH
Class: DCAT-AP classes that are not specified in DCAT-AP CH
Conformance Problem: Currently all mandatory and recommended classes and properties from DCAT-AP are defined in DCAT-AP-CH. Portal managers (“data receivers”) and data publishers (“data providers”) need guidance on what it means for them if a class or property is specified in DCAT-AP but not in DCAT-AP CH: can they use it, will the data portal support it?
Proposal:
Possible consequences on opendata.swiss
Property: schema:image
Class: dcat:Dataset
, dcat:Distribution
Conformance Problem: schema:Image is a custom Swiss property to include thumbnails for datasets and distributions. In DCAT there is a discussion about this topic: w3c/dxwg#1357. In order to stay aligned with DCAT this discussion has to be followed. In case DCAT adds similar vocabulary, it might be necessary to switch to the DCAT vocabulary in the future.
Proposal: Follow the discussion on DCAT development regarding adding a thumbnail property
Class: dcat:Distribution
Conformance Problem:
Details:
dct:coverage
on dcat:Distribution
can be expressed with DCAT-Vocabulary, see here: SEMICeu/DCAT-AP#197 we were told that modeling dataseries with distributions of a dataset is considered an antipattern and the reference above was mentioned in that regard.dcat:DataSeries
(https://www.w3.org/TR/vocab-dcat-3/#Class:Dataset_Series) will be added, that might be the best way to model dataseriesProposal:
Property: dct:issued
Class: dcat:Dataset
Implementation: visiblity of datasets with dct:issued in the future
Conformance to DCAT-AP CH: not conformant to DCAT-AP CH since the usage note on dct:issued says, that datasets with a future relase date should not be published
Details: Currently it is possible to set a datasets visibilty to published regardless of whether the release date of that dateset is in the future or not.
Proposal: Correct implementation on opendata swiss: a dataset's visibility should not be published if the release date is in the future.
Property: dct:accrualPeriodicty
Class: dcat:Dataset
Conformance Problem: DCAT-AP demands the use of the EU vocabulary as a MUST
Details: EU vocabulary: http://publications.europa.eu/resource/authority/frequency
Proposal: restrict the range to the usage of the EU vocabulary and change the implementation on opendata.swiss
Property: dct:coverage
Class:dcat:Dataset
Conformance Problem:
dcat:Dataset
this Swiss custom property is violating the principle that custom properties should not replace available DCAT properties. In this case the DCAT properties dct:spatial
and dct:temporal
are able to completely replace dct:coverage
on dct:Dataset
.Proposal:
dct:coverage
as property of dcat: Dataset
Property: dct:publisher
Class: dcat:Catalog and dcat:Dataset
Conformance Problem: Contradiction in the Definition
Details: Example models this with rdfs:Descripton, but Range says foaf:Agent
Proposal: remove the contradicting example and correct implementation on opendata.swiss
Property: dct:license
Class: dcat:Distribution
Conformance Problem: Currently in DCAT-AP CH dct:license
is optional on dcat:Distribtuion
. This does not conform to DCAT-AP where it is recommend.
Details:
dct:rights
cannot replace dct:license
but just complement it, see #110Proposal:
There are two proposals:
dct:license
should not only be recommended but mandatory on dcat:Distribution
.Property: dct:spatial
Class: dcat:Dataset
Conformance Problem: The requirement level of this property in DCAT-AP CH is "optional" which is weaker than the requirement level in DCAT-AP where it is "recommended".
Details: When overwriting the requirements of a property, DCAT-AP CH must make sure that if metadata conforms to these changed requirements, it will also conform to the original requirements of DCAT-AP. This means in practice that the requirements in DCAT-AP CH may become stricter, but not less strict.
Proposal: Change the requirement level to "recommended"
Property:dct:issued, dct:modified
Class: dcat:Dataset, dcat:Distribution
Implementation: implemented as xsd:datetime
Conformance to DCAT-AP-CH: not conformant with DCAT-AP-CH that demands this to be xsd:date
Conformance to DCAT-AP: conformant witb DCAT-AP that gives the choice of either xsd:date or xsd:datetime
Details: dct:issued
and dct:modified
are demanded as rdfs:Literal
typed as xsd:date
in DCAT-AP CH, but implemented as rdfs:Literal typed as xsd:datetime on opendata.swiss. This implementation does not conform to DCAT-AP CH, but it does conform to DCAT-AP. Therefore it would be an option to adapt the definition of those properties to the DCAT-AP specification in order to solve this non-conformance of opendata.swiss to DCAT-AP CH.
Proposal: change DCAT-AP-CH and allow also for xsd:datetime
Property: dct:availability
Class: dcat:Dataset
Implementation: To be discussed, but to make it clear that it is not a guaranteed sevice, implemented as a CV with a list of brought Availability measures tbd, (e.g. shortTerm (1 year), midTerm (5 years), longTerm (10 years)).
Conformance: For real world applications it is important to quickly understand how long a dataset will stay available to quickly decide if it can be used in a (data-) project. The goal is to provide a data provider statement regard the long-term availability of the provided dataset. This should be made conform to the efforts on the DCAT-AP level, see below.
Property: multiple properties, for example title, description...
Class: -
Ist: DCAT-AP CH does not provide any guidance regarding which language are a must and which not. It just enables multilingualism.
Soll: To increase to usability of data and provide clearer guidance for data publisher at every level it should be clearly stated which language are a must. It can also be diferntiated for different federal levels. For instance: "DE and FR obligatory only for the federal level."
Property: all properties of dcat:Catalog
except dcat:dataset
Class: dcat:Catalog
Implementation: There is no import of the catalog class into opendata.swiss
Conformance to DCAT-AP-CH: Metadata for dcat:Catalog
is currently not imported to opendata.swiss and also not exported from opendata.swiss. Since there are mandatory properties on the catalog class, this hurts DCAT-AP conformance as well
as DCAT-AP CH conformance.
Proposal:opendata.swiss should implement a catalog class and consider to preserve the structure of the imported data catalogs so that this structure is available on export.
Since usage conditions (Nutzungsbedingungen, Lizenzen) are so important, we will propose a controlled vocabulary for dct:license. Important inspiration sources may be http://dcat-ap.de/def/licenses and Chapter 5.4 "Licence vocabularies" from DCAT-AP.
A first draft was elaborated:
https://www.dcat-ap.ch/vocabulary/licenses
Business rules, like for instance "actors of the federal level should only use terms of use", can be defined where needed.
Property: dct:language
Class: dcat:Catalog, dcat:Dataset, dcat:DIstribution
Conformance Problem: DCAT-AP 2.0.0 demands the use of the EU vocabulary for the language, while DCAT-AP-CH demands the ISO 2 letter code
Details: This is the EU vocabulary for the language: https://op.europa.eu/en/web/eu-vocabularies/dataset/-/resource?uri=http://publications.europa.eu/resource/dataset/language
Proposal: Switch to European Language vocabulary
Property: dct:rights, dct:license, dct:format, dcat:mediaType, dct:issued, dct:modified, dcat:landingPage, dcat:byteSize,
Class: dcat:Catalog, dcat:Dataset, dcat:Distribution
Conformance Problem: DCAT-AP CH uses the requirement level "conditional" for specifying the requirements of its classes and properties instead of the requirement level "recommended" that is used and also defined by DCAT-AP. While this may not be an obstacle regarding conformance to DCAT-AP, it makes it harder to compare the requirement levels. Since DCAT-AP CH does not give a definition of "conditional", the reason for that deviation from DCAT-AP terminology remains unclear.
Proposal: Drop support of conditional and replace it with DCAT-AP's "recommended".
Property:dct:identifier
Class:dcat:Dataset
Conformance Problem: The current definition of this property in DCAT-AP CH is very application dependent: it states how the field content should also be used in rdfs:seeAlso and what the application should do to keep the data portal consistent. The Application profile DCAT-AP CH should not include details on how to implement this in a specific data portal (e.g. opendata.swiss), it should focus on providing guidance to Swiss data portals and catalogues.
Proposal: Align to the dcat-ap recomendations. See here: https://joinup.ec.europa.eu/release/dcat-ap-how-use-identifiers-datasets-and-distributions.
Property: dcat:format
Class: dcat:Distribution
Conformance Problem: DCAT-AP makes the use of the EU file type vocabulary mandatory for this property
Details: EU-filetype vocabulary url: http://publications.europa.eu/resource/authority/file-type. It is encouraged to contribute additional file types. A discussion has been started on the file types we currently use on opendata.swiss with [email protected]. There is a plan to enhance their file type vocabulary and to discuss further with the stakeholders about their needs regarding file types
Proposal: Adapt definition and stay in discussion with [email protected].
Property:dct:language
Class:dcat:Catalog
Conformance Problem:dct:language has the requirement level "recommended" in DCAT-AP. In DCAT-AP-CH it is only "optional"
Proposal: upgrade requirement level
To my knowledge there is currently no way to describe attributes of a dataset (e.g. columns of a CSV). This would include the following information (minimal):
On http://data.stadt-zuerich.ch we provide this information on a dataset level (i.e. it does not differ between distributions).
Property: dct:format
Class: dcat:Distribution
Implementation: implemented as an rdfs:Literal, Example dct:format:"CSV"
Conformance: this does not conform to DCAT-AP where a controlled vocabulary should be used: http://publications.europa.eu/resource/authority/filetype
Proposal: wait for DCAT-AP CH to resolve this and then implement it on opendata.swiss
Property: dct:coverage
Class:dcat:Dataset and dcat:Distribtution
Conformance Problem:
dcat:Distribution
there is no DCAT property to replace dct:coverage
. But this is because Distribution properties all relate to resolution, format, size, language, etc. that is to properties that refer to the same content in different physical representations. On opendata.swiss dct:coverage
is used to model dataseries as distributions of a dataset. But this considered bad practice in DCAT-AP, see #112Proposal:
dct:coverage
on dcat:Distribution
and encourage the use of dcat:DataSeries
(https://www.w3.org/TR/vocab-dcat-3/#Class:Dataset_Series) as soon as it is ready to be added to DCAT-AP CH. This has to be done in correspondence with #112Changed Proposal: Watch what happens on DCAT-AP and DCAT regarding the modeling on Series of Data, since this seems to be a topic where there is some movement in the DCAT community.
Please be aware of the following used and automatically created Controlled Vocabularies:
Organisations: https://register.ld.admin.ch/opendataswiss/org
Category: https://register.ld.admin.ch/opendataswiss/category
Licences: https://ld.admin.ch/vocabulary/TermsOfUse
They were created and used because of the lack of any missing list so far.
Please make sure that the CV are accessible through dereferencing in the final version.
See http://dcat-ap.ch/vocabulary/licenses/terms_by, vs. https://ld.admin.ch/vocabulary/TermsOfUse/Provide-the-Source while using the Content Header: Accept: text/turtle
Property:dct:spatial
Class:dcat:Dataset
Conformance Problem
Proposal:
Property: 'dct:issued'
Class: ´dcat:Dataset'
Implementation: in the implementation on opendata.swiss a dataset can be set to published despite the fact that the release date is in the future.
Conformance: This is not conformant to DCAT-AP CH since the usage note on dct:issued says: "In a data portal, the dataset will only be publicly displayed if it has an issued date and if that date is not in the future."
Proposal:
Property: dcat:keyword
, dcat:theme
, dct:spatial
, dct:temporal
Class: dcat:Dataset
Conformance Problem: these properties are optional in DCAT-AP CH but recommended in DCAT-AP
Proposal: make them recommended in DCAT-AP CH
Property: -
Class: dcat:Dataservice
Implementation: Actually is dcat:Dataservice not a part of DCAT-AP CH. This class is becoming increasingly fundamental to reference correctly elements such as SPARQL Endpoints, REST-API, WFS services… Few other countries have already introduced this class. https://docs.dataportal.se/dcat/en/#dcat%3ADataService, https://data.norge.no
Conformance: dcat:DataService is an optional class of DCAT-AP 2.0.1
Proposal: Introducing the class dcat:Dataservice with following properties:
Property:dcat: themeTaxonomy
Class:dcat:Catalog
Conformance Problem:dct:themeTaxonomy has the requirement level "recommended" in DCAT-AP. In DCAT-AP-CH it is only "optional"
Proposal: raise requirement level
Property: dct:availability
Class: dcat:Dataset
Implementation: implemented as a CV with a list tbd, (e.g. Testing, Draft, Published).
Conformance: For real world applications it is important to quickly understand at which maturity level dataset is published to quickly decide if it can be used in a (data-) project. The goal is to provide a data provider statement regard the maturity of the provided dataset. With the goal to also allow to quickly distribute testing and draft dataset for early feedback and to allow an overall agile data modelling process. This is a non-normative extension to the current DCAT.
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.