search

netwerk-digitaal-erfgoed / requirements-datasets

Requirements for datasets
https://netwerk-digitaal-erfgoed.github.io/requirements-datasets/
1 stars 0 forks source link

Requirements schema:DataCatalog cannot hold #52

Open coret opened 2 years ago

coret commented 2 years ago

Our requirements specifiy several properties to be required with schema:DataCatalog. This leads to validation problems, caused by the fact that a Dataset can have an schema:includedInDataCatalog which is of type schema:DataCatalog.

For example:

{
    "@context": "http://schema.org/",
    "@id": "https://data.labs.kadaster.nl/kadaster/bag2",
    "@type": "Dataset",
    ...
    "includedInDataCatalog": {
      "@type": "DataCatalog",
      "@id": "https://data.labs.kadaster.nl",
      "name": "Kadaster Data Science"
    },
    ...
}

This schema:includedInDataCatalog within a schema:Dataset can not be expected to have all datasets of the Datacatalog. Properties like schema:publisher and schema:name are less likely to be present in schema:includedInDataCatalog.

Prososal: change the cardinality and usage of properties to make them optional (after which this should also be change in the SHACL).

4.6.4. DataCatalog attributes

Property | Description | Cardinality | Usage -- | -- | -- | -- @id | The HTTP [IRI] of the data catalog. | 1 | Required schema:name|The name of the data catalog. | **0..1** | **Recommended** schema:description|A description of the data catalog. | 0..1 | Recommended schema:publish|The publisher of the data catalog. | **0..1** | **Recommended** scheme:dataset|The datasets that are contained in the data catalog. | **0**..n | **Recommended**
ddeboer commented 2 years ago

If the SHACL needs to change, only change that but leave the requirements as they are to invite providers to submit full data catalog descriptions.

coret commented 1 year ago

Additional cases where the strict requirements hinder validation: