openDS term data schema #196

sharifX · 2024-09-18T08:52:58Z

A JSON schema to describe the openDS terms as FDO.
The first version of this schema is focused on creating the basic structure of the object. Feedback from RFC will be incorporated in different stages. There are a few decisions regarding RDF, JSON-LD we have to make.

A JSON schema to describe the openDS terms as FDO.

samleeflang

Ok I think the start is there but I would make some changes:

Don't mix schema.org and dcterms terms, there is a lot of overlap but I would say we pick one and stick with it
Keep to naming convention
BelongToClass probably needs to change but that impacts some other fields as well

data-model/fdo-type/term/term.json

- Used dcterms properties to be consistent with the rest of the digital objects. Removed schema.org properties. - Fixed naming style to be consistent with the rest. - Followed ISO standards and examples for dcterms:created and dcterms:modified. - Removed contributor, seemed not needed for now. We can use Creator as an Agent object. This should be sufficient for now. - Removed mapping. As mentioned in the pull request when we add a mapping to other standards, we would need to update the term, which might get tricky. - Added const spdx URl for license. - To handle "repeatable" and "required" properties, using existing terms is better, so we are using "owl:cardinality", "owl:minCardinality", and "owl:maxCardinality". - If a term can be required for one class but optional for another, this introduces complexity because JSON Schema doesn't easily support conditional logic like this. This might need to addressed in the application layer.

samleeflang

Might be best to present this in the exploration meeting, I have difficulty grasping how this would function in the landscape? What will the identifiers look like, do we need changes to the terms (for example add version)? How does this fit with semantic best practise (we use owl terms but don't implement owl)?

data-model/fdo-type/term/term.json

samleeflang · 2024-10-15T12:17:21Z

data-model/fdo-type/term/term.json

+    "dcterms:creator": {
+      "type": "object",
+      "description": "Contains information about the creator/proposer/requester of the term. This is an ods:Agent object",
+      "$ref": "https://schemas.dissco.tech/schemas/fdo-type/shared-model/0.4.0/agent.json",


I am not sure about the use of a openDS object here. The example would look a bit differently as the orcid should also be there as an identifier.

data-model/fdo-type/term/term.json

samleeflang · 2024-10-15T12:23:36Z

data-model/fdo-type/term/term.json

+      "const": "https://spdx.org/licenses/CC0-1.0.json",
+      "description": "The licence for the term, required to be always public domain"
+    },
+    "ods:belongsToClass": {


Wondering if rdfs:range does not cover this? This doesn't feel like we should be the ones creating this term

check if rdfs:range. can this do array?

we are using rdfs:range. updated.

data-model/fdo-type/term/term.json

changes in this version: - updated Regex for the Handle - removed rdf:Type - used ods:hasAgent instead of dcterms:creator - added skos:note and skos:example - removed owl cardinality. we use json schema required for this - added hasVersion for more provenance metadata - used rdfs:range instead of new term belongsToClass - added langiage in prefLabel

sharifX · 2024-12-17T09:45:23Z

updated schema
changes in this version:

updated Regex for the Handle
removed rdf:Type
used ods:hasAgent instead of dcterms:creator
added skos:note and skos:example
removed owl cardinality. we use json schema required for this
added hasVersion for more provenance metadata
used rdfs:range instead of new term belongsToClass
added language in prefLabel

samleeflang

Couple comments. General points:

Can a term reference itself? We use odsTerms here which are described by this object. So the object is using a term which is defined by itself. Not sure if that is an issue or not.
I think we need a lot more required attributes

data-model/fdo-type/term/term.json

@type

Changes (based on the PR and meeting feedback) - updated required properties - added hasAgents (this is another object we will call from this) - removed dcterms:hasVersion. Even though this is semantically diff than ods:version we do not need this now. - added "@type" - Regarding self reference. yes, that was answered in the PR comment. Yes, this is common in RDF/ontology and data typing world. self referencing/recursive definition. For example in RDF schema (https://www.w3.org/2000/01/rdf-schema) rdfs:Class is the class of all classes, and it is itself an instance of rdfs:Class.

Added optional ods:hasControlledValues array to define controlled vocabulary values that a term accepts (e.g., topicDiscipline can have values: Botany, Geology, etc.). Each controlled value gets its own PID following pattern hdl:20.500.1025/term/value. This complements skos:inScheme usage and maintains FAIR/FDO compliance while avoiding misuse of skos:member outside of collections. skos:narrower: Implies hierarchical relationships where the controlled values are "narrower concepts" of the term itself. But that's not quite right - "Botany" isn't a narrower concept of "topicDiscipline", it's a value that topicDiscipline can take. skos:related: Too vague and doesn't capture the specific "this term accepts these values" relationship

samleeflang

Have some comments, mainly about the examples.

samleeflang · 2025-08-20T13:17:07Z

data-model/fdo-type/term/examples/ods:topicDiscipline-example.json

+  "ods:fdoType": "https://doi.org/21.T11148/bbad8c4e101e8af01115",
+  "ods:status": "ods:Active",
+  "ods:version": 1,
+  "ods:hasAgents": [],


This should have a creator and/or modifier

samleeflang · 2025-08-20T13:19:37Z

data-model/fdo-type/term/examples/ods:topicDiscipline-example.json

+{
+  "@id": "https://hdl.handle.net/20.500.1025/topicDiscipline",
+  "@type": "ods:Term",
+  "ods:fdoType": "https://doi.org/21.T11148/bbad8c4e101e8af01115",


This points to the digital media fdoType record. Is there already a FDO type for this otherwise we need to add it later and don't forget to update the examples

created a new type 21.T11148/770bce19da07e2e1106c

samleeflang · 2025-08-20T13:20:12Z

data-model/fdo-type/term/examples/ods:topicDiscipline-example.json

+  "skos:definition": "Add def here.",
+  "dcterms:created": "2025-01-15T09:00:00.000Z",
+  "dcterms:modified": "2025-01-15T09:00:00.000Z",
+   "rdfs:range": [


Hmm this points to itself at the moment, it should point to DigitalSpecimen

samleeflang · 2025-08-20T13:21:43Z

data-model/fdo-type/term/examples/ods:topicDiscipline-example.json

+  "ods:hasAgents": [],
+  "dcterms:title": "ods:topicDiscipline",
+  "skos:prefLabel": "Topic Discipline",
+  "skos:definition": "Add def here.",


Maybe take over the existing definition here:
"Overarching classification of the scientific discipline to which the objects within the collection belong or are related"

samleeflang · 2025-08-20T13:22:22Z

data-model/fdo-type/term/examples/ods:topicDiscipline-example.json

+  "skos:prefLabel": "Topic Discipline",
+  "skos:definition": "Add def here.",
+  "dcterms:created": "2025-01-15T09:00:00.000Z",
+  "dcterms:modified": "2025-01-15T09:00:00.000Z",


Would be better to have a time difference with dcterms:created. If we have modified in here we should also have an agent with a role Modifier

samleeflang · 2025-08-20T13:23:10Z

data-model/fdo-type/term/examples/ods:physicalSpecimenID-example.json

+  "skos:note": "Used in multiple schemas including the Digital Specimen FDO.",
+  "skos:example": "https://data.biodiversitydata.nl/naturalis/specimen/RGM.234626",
+  "dcterms:created": "2025-07-30T10:00:00Z",
+  "dcterms:modified": "2025-07-30T10:00:00Z",


If we have a modified I would expect a modifier

samleeflang · 2025-08-20T13:25:20Z

data-model/fdo-type/term/term.json

+    },
+    "ods:fdoType": {
+      "type": "string",
+      "description": "The DOI to the FDO type of the object",


The FDO types are currently handles so this should be a handle string. See https://schemas.dissco.tech/schemas/fdo-type/virtual-collection/0.1.0/virtual-collection.json

samleeflang · 2025-08-20T13:25:27Z

data-model/fdo-type/term/examples/ods:physicalSpecimenID-example.json

+{
+  "@id": "https://hdl.handle.net/20.500.1025/physicalSpecimenID",
+  "@type": "ods:Term",
+  "ods:fdoType": "https://doi.org/21.T11148/bbad8c4e101e8af01115",


Needs to be updated when we have an fdoType handle (also better to put this as a handle as it is not a DOI)

Fixed hasAgents, added modifier role added correct fdoType PID.

samleeflang

Check the comments, but there are still some basic issue with this PR.
Please validate that the properties you use to describe are valid JSON Schema properties. You can just check the standard https://json-schema.org/.
Validate the examples against the schema.
This can easily be done through something like https://www.jsonschemavalidator.net/
You will immediately see there are issue with the validation of the examples.

samleeflang · 2025-10-29T07:33:39Z

data-model/fdo-type/term/term.json

+        "type": "string"
+      }
+    },
+    "skos:member": {


Formatting is off here and on other places in this file. Clear formatting helps with readability.

ah yes, this can be removed. ods:hasControlledvalue does the same purpose justification is that the proper use of skos:member is related to skos:Collections. we are not using that. i added some notes in DEX-106.

and formatting fixed.

samleeflang · 2025-10-29T07:50:26Z

data-model/fdo-type/term/term.json

+      "examples": [
+        "Digital Specimen"
+      ],
+      "@language": {


This isn't a valid JsonSchema property

I will remove this.

data-model/fdo-type/term/term.json

samleeflang · 2025-10-29T07:55:52Z

data-model/fdo-type/term/term.json

+        "type": "string"
+      }
+    },
+    "skos:member": {


I don't understand how this works and I don't see it used in the examples. Wouldn't all these terms be of the dissco vocabulary?

this can be removed. ods:hasControlledvalue does the same purpose justification is that the proper use of skos:member is related to skos:Collections. we are not using that. i added some notes in DEX-106. and yes. each term will be a dissco vocab. ods:topicDiscipline example has this.

samleeflang · 2025-10-29T07:57:59Z

data-model/fdo-type/term/term.json

+    "skos:definition",
+    "ods:hasAgents",
+    "ods:status",  
+    "rdfs:range"


Would range always be there? If I got a ods:DigitalSpecimen which is the class than it won't have a range correct?

yes, this does not need to be required.

samleeflang · 2025-10-29T08:07:57Z

data-model/fdo-type/term/examples/ods:physicalSpecimenID-example.json

+    "https://hdl.handle.net/20.500.1025/DigitalSpecimen"
+  ],
+  "owl:deprecated": false,
+  "ods:hasAgents": [


Agent lacks the ods:hasIdentifier object for the orchid see any of the other examples:

"dcterms:creator": { "@id": "https://orcid.org/0000-0002-5669-2769", "@type": "schema:Person", "schema:name": "John Doe", "ods:hasIdentifiers": [ { "@id": "https://orcid.org/0000-0002-5669-2769", "@type": "ods:Identifier", "dcterms:title": "ORCID", "dcterms:type": "URL", "dcterms:identifier": "https://orcid.org/0000-0002-5669-2769", "ods:isPartOfLabel": false, "ods:gupriLevel": "GloballyUniqueStablePersistentResolvable", "ods:identifierStatus": "Preferred" } ] },

I will update the example.

samleeflang · 2025-10-29T08:08:37Z

data-model/fdo-type/term/examples/ods:topicDiscipline-example.json

+        {
+          "@type": "schema:Role",
+          "schema:roleName": "modifier",
+          "schema:startDate": "2025-01-17T09:00:00.000Z",


We haven't really used the start/endDate of role yet and I am not sure how it will work. I would remove them here. Remove these and add the hasIdentifier part

samleeflang · 2025-10-29T08:09:55Z

data-model/fdo-type/term/examples/ods:physicalSpecimenID-example.json

+  "skos:definition": "The physical specimen identifier of the object. The main identifier of the object is used (which can be indicated by the user through the DataMapping). If the ods:physicalSpecimenIDType is Resolvable or Global, this identifier is the same as the `ods:normalisedPhysicalSpecimenID`.",
+  "skos:note": "Used in multiple schemas including the Digital Specimen FDO.",
+  "skos:example": "https://data.biodiversitydata.nl/naturalis/specimen/RGM.234626",
+  "dcterms:created": "2025-07-30T10:00:00Z",


This object lacks the required field "dcterms:modified"

samleeflang · 2025-10-29T08:10:06Z

data-model/fdo-type/term/examples/ods:topicDiscipline-example.json

+{
+  "@id": "https://hdl.handle.net/20.500.1025/topicDiscipline",
+  "@type": "ods:Term",
+  "ods:fdoType": "https://doi.org/21.T11148/770bce19da07e2e1106c",


This should be a handle

samleeflang · 2025-10-29T08:13:27Z

data-model/fdo-type/term/term.json

+      "description": "The Handle to the FDO type of the object",
+      "pattern": "^https:\/\/hdl\\.handle\\.net\/[\\w.]+\/(.){3}-(.){3}-(.){3}",
+      "examples": [
+        "https://hdl.handle.net/21.T11148/2ac65a933b7a0361b651"


This is still the VirtualCollection, better to point to the FDOType for the Term here

- Updated regex patterns for FDO type - removed skos:member as this is same as hasControlledvalue - removed rdfs:range from required - added skos:example in required - removed @langauge as this is not valid JSON

- updated fdoType link to the correct Handle - updated regex - updated hasAgents to have hasIdentifier.

openDS term data schema

3636fac

A JSON schema to describe the openDS terms as FDO.

sharifX requested review from samleeflang and wouteraddink September 18, 2024 08:53

samleeflang reviewed Sep 20, 2024

View reviewed changes

wouteraddink requested changes Oct 1, 2024

View reviewed changes

data-model/fdo-type/term/term.json Outdated Show resolved Hide resolved

data-model/fdo-type/term/term.json Outdated Show resolved Hide resolved

data-model/fdo-type/term/term.json Outdated Show resolved Hide resolved

sharifX requested review from samleeflang and wouteraddink October 14, 2024 11:17

samleeflang reviewed Oct 15, 2024

View reviewed changes

sharifX requested a review from samleeflang December 17, 2024 09:44

samleeflang reviewed Jul 4, 2025

View reviewed changes

sharifX added 2 commits July 30, 2025 11:07

An example based on the latest Jul 2025 schema version

525cfde

sharifX requested a review from samleeflang July 30, 2025 09:13

sharifX added 3 commits July 31, 2025 14:19

fix rdfs range

41c03a3

Create ods:topicDiscipline-example.json

5ee1226

samleeflang reviewed Aug 20, 2025

View reviewed changes

sharifX added 3 commits October 28, 2025 13:25

Update topicDiscipline example

9283d26

Fixed hasAgents, added modifier role added correct fdoType PID.

Update fdoType URL in physicalSpecimenID example

7254507

Update ods:fdoType description and pattern

44a3cab

sharifX requested a review from samleeflang October 28, 2025 12:28

samleeflang reviewed Oct 29, 2025

View reviewed changes

sharifX added 2 commits November 3, 2025 10:59

update several things based on PR

5bff3b8

- Updated regex patterns for FDO type - removed skos:member as this is same as hasControlledvalue - removed rdfs:range from required - added skos:example in required - removed @langauge as this is not valid JSON

fixed fdoType link regex, and hasIdentifier

d6c71de

- updated fdoType link to the correct Handle - updated regex - updated hasAgents to have hasIdentifier.

openDS term data schema #196

Are you sure you want to change the base?

openDS term data schema #196

Uh oh!

Conversation

sharifX commented Sep 18, 2024

Uh oh!

samleeflang left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

samleeflang left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

sharifX commented Dec 17, 2024

Uh oh!

samleeflang left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

samleeflang left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

samleeflang left a comment

Choose a reason for hiding this comment