Profile SHACL shape #84

VirginiaBalseiro · 2023-01-17T14:42:13Z

Based off: https://github.com/solid/webid-profile/blob/main/shapes/profile-shapes.ttl

csarven · 2023-01-17T14:55:14Z

shapes/profile-shape.ttl

+                            :zeroOrMorePath rdf:rest ]
+                   rdf:first ) ] ];
+            :path schema:knowsLanguage;
+            :targetClass rdf:Collection ] ],


Not sure if this is correct per SHACL but as I understand schemaorg, expected range is Text or Language.

Suggested change

:targetClass rdf:Collection ] ],

:targetClass schema:Text, schema:Language ] ],

Agreed (except that :targetClass should be sh:class). schema:knowsLanguage can just be in the profile multiple times, just as schema:skills, so no need for a complex rdf:Collection structure.

The code uses the structure to allow you to put your languages in order. I think that is quite important in practice. People's ability in different language tends to vary a lot.

Suggestions above seem good and useful.

I'll just leave a reminder about the following. Mentioned the Cognitive Characteristics Ontology ( https://smiy.sourceforge.net/cco/spec/cognitivecharacteristics.html ) elsewhere but it may serve as useful input to towards providing a solution in particular for example ( #52 ):

Factors specific to physical, physiological, genetic, mental, economic, cultural, social, or behavioural identity.

In my https://csarven.ca/cv , I say I have cco:skill and refer to the human-languages in ESCO. I didn't bother to order or put weights (and maybe that's overkill.. not sure.). The details seems useful for CV/Resume-like cases but not necessarily generic Profile info? Could go either way.. especially if the application can simply adapt the UI to use the user's preferred or most skilled language first, then try others (or even give option to change.) So, ordering languages can be simple to input and consume.

Noting the discussion at https://github.com/solid/webid-profile/blob/main/meetings/2023-02-21.md#shapes-pr . Pardon me for the initial suggestion to use schema's Text or Language. I was merely going with the definition of schema:knowsLanguage that appears to use them as rangeIncludes.

I agree that URIs would be preferable to strings. Ditto Collection (especially if that's intended to be interpreted as preferred order for use).

Feel free to revert or change the shape as you see fit.

Re Collections: the semantics of schema:knowsLanguage do not allow for that; the range consists of schema:Language and schema:text. If we want order/weight, I suggest we indeed go with something like CCO as @csarven mentioned, even though this might make profiles quite complex for the human reader.

timea-solid

I will just ask: should this not be a .shacl file?

woutermont

Great job already!

I proposed some structural simplifications where I think it became a bit to complicated, and I tried to clarify the admittedly weird role of sh:targetClass.

shapes/profile-shape.ttl

woutermont · 2023-01-17T15:13:13Z

shapes/profile-shape.ttl

+                :name "Organization"@en,
+                       "Organization"@fr;


This is nice!

woutermont · 2023-01-17T15:18:56Z

shapes/profile-shape.ttl

+                    [
+                :name "occupation";
+                :path org:role ],
+                    [
+                :datatype xsd:string;
+                :path vcard:role ] ];


One is missing a name, the other a datatype.

If these happen to be the same, we can remove redundancy by using a sh:or:

Suggested change

[

:name "occupation";

:path org:role ],

[

:datatype xsd:string;

:path vcard:role ] ];

[

sh:datatype xsd:string;

sh:name "Occupation"@en;

sh:or ( [ sh:path org:role ] [ sh:path vcard:role ] ) ];

shapes/profile-shape.ttl

woutermont · 2023-01-17T15:47:15Z

shapes/profile-shape.ttl

+                            :zeroOrMorePath rdf:rest ]
+                   rdf:first ) ] ];
+            :path schema:knowsLanguage;
+            :targetClass rdf:Collection ] ],


Agreed (except that :targetClass should be sh:class). schema:knowsLanguage can just be in the profile multiple times, just as schema:skills, so no need for a complex rdf:Collection structure.

woutermont · 2023-01-17T15:57:31Z

Apart from what's arleady in the document, I think it might also be interesting to add a small NodeShape for the PersonalProfileDocument entry linking to WebID Document to the Person. Something along the following lines.

:ProfileDocShape     a sh:NodeShape;
    dct:title "ShaCL shape for a PersonalProfileDocument";
    sh:targetClass foaf:PersonalProfileDocument;
    sh:property [
        sh:path foaf:primaryTopic;
        sh:node :ProfileShape;
    ].

jeswr · 2023-01-17T21:38:56Z

I will just ask: should this not be a .shacl file?

@timea .shacl is reserved for SHACL compact syntax; the file in this PR uses Turtle syntax.

…uage

jeswr · 2023-04-01T00:13:35Z

shapes/profile-shape.ttl

+    sh:property 
+            [
+        sh:datatype xsd:string;
+        sh:name "his/hers/theirs...";


A large amount of the sh:names seem better suited to sh:descriptions

Suggested change

sh:name "his/hers/theirs...";

sh:name "Preferred Relative Pronoun";

sh:description "his/hers/theirs...";

Also, these are not and should not be referred to as the person's "preferred" pronouns; rather, they are simply the person's pronouns.

jeswr · 2023-04-01T00:17:46Z

shapes/profile-shape.ttl

+    sh:property 
+            [
+        sh:datatype xsd:string;
+        sh:name "his/hers/theirs...";
+        sh:path solid:preferredRelativePronoun ],
+            [
+        sh:datatype xsd:string;
+        sh:name "him/her/them...";
+        sh:path solid:preferredObjectPronoun ],
+            [
+        sh:datatype xsd:string;
+        sh:name "he/she/they...";
+        sh:path solid:preferredSubjectPronoun ],


I'd suggest making this all part of a sh:group to assist form tools.

My personal opinion is that we should most shapes for client-client standards in Solid should be required to add these non-validating characteristics so that it is easier for apps to generate nice looking forms for users to supplement and/or fix data that is required for their application from a generic form editor.

jeswr · 2023-04-01T00:25:17Z

shapes/profile-shape.ttl

+[
+        sh:name "Set type";
+        sh:path rdf:type;
+        sh:targetValue schema:Person 


Do you mean sh:hasValue?

jeswr · 2023-04-01T00:28:58Z

shapes/profile-shape.ttl

+            [
+        sh:min 1;
+        sh:or (
+[
+        sh:name "Set type";
+        sh:path rdf:type;
+        sh:targetValue schema:Person 
+]
+[
+        sh:name "Set type";
+        sh:path rdf:type;
+        sh:targetValue foaf:Person ]
+[
+        sh:name "Set type";
+        sh:path rdf:type;
+        sh:targetValue vcard:Individual 
+])],


This should be able to just be

sh:path rdf:type; sh:qualifiedMinCount 1 ; sh:qualifiedValueShape [ sh:in ( schema:Person foaf:Person vcard:Individual ) ]

jeswr · 2023-04-01T00:30:17Z

shapes/profile-shape.ttl

+                sh:name "Organization"@en,
+                       "Organization"@fr;
+                sh:path org:organization ],


Perhaps set the sh:nodeKind as sh:blankNodeOrIRI

pmcb55 · 2023-11-17T10:49:19Z

I will just ask: should this not be a .shacl file?

@timea .shacl is reserved for SHACL compact syntax; the file in this PR uses Turtle syntax.

@jeswr - just being pedantic, but that draft SHACL compact syntax spec you reference states that the reserved file extensions are either .shc and .shaclc (i.e., with a trailing 'c'). But yeah, your point is still valid though - i.e., this file is indeed using the Turtle syntax, and so .ttl is what I'd expect - but @timea-solid, have you ever come across a use of .shacl as a file extension (I can't find any from a quick Google search)...?

VirginiaBalseiro mentioned this pull request Jan 17, 2023

Add PB profile shape for discussion #79

Closed

csarven reviewed Jan 17, 2023

View reviewed changes

timea-solid reviewed Jan 17, 2023

View reviewed changes

woutermont suggested changes Jan 17, 2023

View reviewed changes

VirginiaBalseiro marked this pull request as ready for review January 24, 2023 07:31

VirginiaBalseiro changed the title ~~WIP: profile SHACL shape~~ Profile SHACL shape Jan 24, 2023

VirginiaBalseiro and others added 3 commits March 31, 2023 14:19

WIP: profile SHACL shape

16783bd

Implement suggestions from review

86e28d5

reintroduce Collection as a possible targetClass for schema:knowsLang…

9141123

…uage

VirginiaBalseiro force-pushed the profile-shacl-shape branch from 0a876cc to 9141123 Compare March 31, 2023 12:26

VirginiaBalseiro requested review from timbl, woutermont, csarven and timea-solid March 31, 2023 12:27

jeswr reviewed Apr 1, 2023

View reviewed changes

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Profile SHACL shape #84

Profile SHACL shape #84

VirginiaBalseiro commented Jan 17, 2023 •

edited

Loading

csarven Jan 17, 2023

woutermont Jan 17, 2023

timbl Jan 25, 2023

csarven Feb 20, 2023

csarven Feb 21, 2023

woutermont Jul 8, 2023

timea-solid left a comment

woutermont left a comment

woutermont Jan 17, 2023

woutermont Jan 17, 2023

woutermont Jan 17, 2023

woutermont Jan 17, 2023

woutermont commented Jan 17, 2023

jeswr commented Jan 17, 2023

jeswr Apr 1, 2023 •

edited

Loading

TallTed Apr 3, 2023

jeswr Apr 1, 2023

jeswr Apr 1, 2023

jeswr Apr 1, 2023

jeswr Apr 1, 2023 •

edited

Loading

pmcb55 commented Nov 17, 2023

	:targetClass rdf:Collection ] ],
	:targetClass schema:Text, schema:Language ] ],

	sh:name "his/hers/theirs...";
	sh:name "Preferred Relative Pronoun";
	sh:description "his/hers/theirs...";

Profile SHACL shape #84

Are you sure you want to change the base?

Profile SHACL shape #84

Conversation

VirginiaBalseiro commented Jan 17, 2023 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

timea-solid left a comment

Choose a reason for hiding this comment

woutermont left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

woutermont commented Jan 17, 2023

jeswr commented Jan 17, 2023

jeswr Apr 1, 2023 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

jeswr Apr 1, 2023 • edited Loading

Choose a reason for hiding this comment

pmcb55 commented Nov 17, 2023

VirginiaBalseiro commented Jan 17, 2023 •

edited

Loading

jeswr Apr 1, 2023 •

edited

Loading

jeswr Apr 1, 2023 •

edited

Loading