Update RDF-semantics with the liberal baseline RDF and RDFS semantics #74

franconi · 2025-01-27T18:16:41Z

Update RDF-semantics with the liberal baseline RDF and RDFS semantics

Summary of changes:

fixed definition of instance in Section 4
definition of "appears in" in Section 4
added rdf:reifies in RDF vocabulary (Section 8)
modified entailment pattern rdfD2 (Section 8.1.1)
added rdfs:Proposition in RDFS vocabulary (Section 9)
added RDFS semantic condition for rdfs:Proposition (Section 9)
added axiomatic triple about range of rdf:reifies (Section 9)
modified entailment pattern rdfs4 (Section 9.2.1)
addded entailment pattern rdfs14 (Section 9.2.1)
added example of incomplete rdfs:Proposition entailment (Appendix A)
added complete entailment pattern Grdfs14 for generalised RDF (Appendix A)

Open issue:

suggest a better way to write entailment pattern rdfs14 (Section 9.2.1)

TODO

check Section 6: Skolemisation
check Appendix C: Reification
check whole text for consistency

Preview | Diff

…or rdfs:Proposition

TallTed

Editorial. Small language fixes.

spec/index.html

gkellogg · 2025-01-27T22:02:03Z

There are several markup problems that generate Echidna errors:

Error: ROR] Couldn't find ":" in this document or other cited documents: [horst04], [horst05], [owl2-profiles], and [rdf12-concepts].
Error: ROR] Couldn't find "****" in this document or other cited documents: [horst04], [horst05], [owl2-profiles], and [rdf12-concepts].
Error: ROR] Couldn't find "****" in this document or other cited documents: [rdf12-concepts].
Error: ROR] Couldn't find "Finite interpretations To keep the exposition simple, the RDF semantics has been phrased in a way which requires interpretations to be larger than absolutely necessary. For example, all interpretations are required to interpret the whole IRI vocabulary, and the universes of all D-interpretations where D contains xsd:string must contain all possible strings and therefore be infinite. This appendix sketches, without proof, how to re-state the semantics using smaller semantic structures without changing any entailments." in this document or other cited documents: [rdf12-concepts].
Error: ROR] Couldn't find "Basically, it is only necessary for an interpretation structure to interpret the" in this document or other cited documents: [rdf12-concepts].
Error: ROR] Error in handler for topic "toc": Cannot read properties of null (reading 'textContent')

Likely, there's some missing closing tag that causes the parser to blow up. It's worth validating the HTML before pushing.

franconi · 2025-01-28T09:36:48Z

It's worth validating the HTML before pushing.

Thanks. I did it, and I have committed now an error-free html.

Co-authored-by: Ted Thibodeau Jr <[email protected]>

spec/index.html

pfps

Looks good enough aside from the changes I mentioned.

spec/index.html

pchampin

generally LGTM

spec/index.html

pchampin · 2025-01-29T14:07:29Z

spec/index.html

+        <tbody>
+          <tr>
+            <th></th>
+            <th><strong>if the triple or triple term <a>appears in</a> S</strong></th>


Suggested change

<th>if the triple or triple term <a>appears in</a> S</th>

<th>if the triple <a>appears in</a> S</th>

"or triple term" is redundant. The definition of "appears in" for triples covers it.

I'm not opposed to keeping it, but I slightly prefer to remove it.

I believe it would be wrong to assume that the check should be done only for triples.

pchampin · 2025-01-29T14:19:07Z

spec/index.html

+        <tbody>
+          <tr>
+            <th ></th>
+            <th >If the triple or triple term <a>appears in</a> S:</th>


Suggested change

<th >If the triple or triple term <a>appears in</a> S:</th>

<th >If the triple <a>appears in</a> S:</th>

same as above

As above :-), I believe it would be wrong to assume that the check should be done only for triples.

pchampin · 2025-01-29T14:24:20Z

spec/index.html

+          <tr >
+            <td class="othertable"><dfn>rdfs14</dfn></td>
+            <td class="othertable">xxx yyy &lt;&lt;(aaa bbb ccc)>> <code></code></td>
+            <td class="othertable">the graph S with the appearance replaced with <br/>


Suggested change

<td class="othertable">the graph S with the appearance replaced with 

<td class="othertable">

No need for that. RDF is monotonic, so any inferred triple can coexist with the original triples.
In fact, this sentence may give the wrong impression that the original triples must be removed.

Sure, but we need to ADD the triples with the bnode substitution.
On the other hand, we should find a better way to say it.

Oh, sorry, now I get it!
The triples on the right are only entailed as is IF the triples on the right are asserted.
But since the rule says they must "appear", then we need to replace their "appearance"...

It is not clear, however, than only the 1st of the 2 triples needs to be "injected". The other one is inferre asserted.

Also, a number of the triples in the rules are missing the trailing period (.).

spec/index.html

pchampin · 2025-01-29T14:43:02Z

spec/index.html

+
+      <tr>
+        <th> </th>
+        <th>RDFS entails</th>
+      </tr>
+
+      <tr>
+        <td class="othertable"><code>ex:a ex:b ex:c .</code></td>
+       <td class="othertable"><code>&lt;&lt;(ex:a ex:b ex:c)>> rdf:type rdfs:Proposition .</code> </td>
+      </tr>


This addition is present in the papers referenced at the top of this section [HORST04,HORST05]. I think the introductory sentence of that section should be explicit about the fact that the cited work was extended in RDF 1.2 with triple terms.

Wishful thinking. We have NOT extended the cited work in RDF 1.2 with triple terms. It should not be very hard to deal with this case, but I leave it to @pfps .

rat10 · 2025-01-29T15:11:13Z

RDFS axiomatic triples. says rdf:reifies rdf:range rdfs:Proposition . That must be rdfs:range.

spec/index.html

Co-authored-by: Pierre-Antoine Champin <[email protected]>

Co-authored-by: Niklas Lindström <[email protected]>

franconi · 2025-01-29T16:14:48Z

@pfps: I think I resolved all your reviews, but you still appear in the final summary as having a requested change. I don't understand why.

pfps · 2025-01-29T16:18:31Z

It's a github thing. I have to sign off somehow.

Done.

pfps

Looks good now

franconi · 2025-01-29T17:29:18Z

@doerthe: can you please check whether you like how rdfs14 is written? Maybe propose a better way? Or maybe could be fine if we add an example later?

pchampin · 2025-01-29T18:05:44Z

spec/index.html

+      exist x,y,z such that RE(x,z,y)=r, or exists x such that &lt; x,r &gt; is in IEXT([I+A](<code>rdf:reifies</code>)), <br/> then &lt; r,[I+A](<code>rdfs:Proposition</code>)&gt; is in IEXT([I+A](<code>rdf:type</code>))</p></td>
+    </tr>


I think that [I+A] should be replaced with I here. [I+A] is never mentioned in the semantic conditions of RDF-Entailment or RDFS-Entailment.

TallTed · 2025-01-29T19:40:42Z

spec/index.html

+    An RDF term <dfn>appears in</dfn> a graph G if it is in the set TM(G), defined inductively as follows: (1) for each triple in the graph G, its subject, predicate, and object are in TM(G); (2) for each triple term in TH(G), its subject, predicate, and object are in TM(G).<br/>
+    A triple <b><i>appears in</i></b> a graph G if it is in G or if the triple term composed by its subject, predicate, and object appears in G.


I've not changed any text here, only white space in the source, where shorter lines are far easier to review. Also, if it's worth  , it's usually worth  . Also also, I'm wondering why the (1)(2) don't warrant   or <ol>.

Suggested change

An RDF term <dfn>appears in</dfn> a graph G if it is in the set TM(G), defined inductively as follows: (1) for each triple in the graph G, its subject, predicate, and object are in TM(G); (2) for each triple term in TH(G), its subject, predicate, and object are in TM(G). 

A triple appears in a graph G if it is in G or if the triple term composed by its subject, predicate, and object appears in G.

An RDF term <dfn>appears in</dfn> a graph G if it is in the set TM(G),

defined inductively as follows:

(1) for each triple in the graph G,

its subject, predicate, and object are in TM(G);

(2) for each triple term in TH(G),

its subject, predicate, and object are in TM(G).

 

A triple appears in a graph G

if it is in G or

if the triple term composed by its subject, predicate, and object appears in G.

TallTed · 2025-01-29T19:44:07Z

spec/index.html

@@ -1206,6 +1231,10 @@ <h2>RDFS Interpretations</h2>
      x is in ICEXT(I(<code>rdfs:Datatype</code>)) then <span >&lt; x,
      I(<code>rdfs:Literal</code>) &gt; is in IEXT(I(<code>rdfs:subClassOf</code>))</span></p></td>
    </tr>
+    <tr>
+    <td class="semantictable" id="rdfssemcond11"><p>If
+      exist x,y,z such that RE(x,z,y)=r, or exists x such that &lt; x,r &gt; is in IEXT([I+A](<code>rdf:reifies</code>)), <br/> then &lt; r,[I+A](<code>rdfs:Proposition</code>)&gt; is in IEXT([I+A](<code>rdf:type</code>))</p></td>


Line lengths, again. I'm not changing it, but I think exist in the first line below should be exists. And, this is the rare case where   is preferred over  .

Suggested change

exist x,y,z such that RE(x,z,y)=r, or exists x such that < x,r > is in IEXT([I+A](<code>rdf:reifies</code>)), then < r,[I+A](<code>rdfs:Proposition</code>)> is in IEXT([I+A](<code>rdf:type</code>))</td>

exist x,y,z such that RE(x,z,y)=r, or

exists x such that < x,r >

is in IEXT([I+A](<code>rdf:reifies</code>)),

 

then < r,[I+A](<code>rdfs:Proposition</code>)>

is in IEXT([I+A](<code>rdf:type</code>))</td>

TallTed · 2025-01-29T19:48:46Z

spec/index.html

@@ -1639,6 +1698,28 @@ <h2>Entailment rules (Informative)</h2>

  <p>As noted earlier, detecting datatype entailment for larger sets of datatype IRIs
    requires attention to idiosyncratic properties of the particular datatypes.</p>
+
+  <p>
+  The complete entailment pattern for generalized RDF with [=symmetric RDF triple=]s, considering that, according to the semantics, the denotation of triple terms should be of type <code>rdfs:Proposition</code>, is the following:


Line length again. =]s should not exist. Any terms which are currently treated this way should get the plural added to the synonyms of the singular, and the s=] should become s=]. (Similar treatment also applies to other plural forms.)

Suggested change

The complete entailment pattern for generalized RDF with [=symmetric RDF triple=]s, considering that, according to the semantics, the denotation of triple terms should be of type <code>rdfs:Proposition</code>, is the following:

The complete entailment pattern for generalized RDF with [=symmetric RDF triples=],

considering that, according to the semantics, the denotation of triple terms should

be of type <code>rdfs:Proposition</code>, is the following:

franconi added 3 commits January 27, 2025 17:42

First version of liberal semantics - it lack the entailment pattern f…

4300b64

…or rdfs:Proposition

Missing fix to rdfs14

61aa733

liberal baseline RDF and RDFS semantics

553dfc9

franconi self-assigned this Jan 27, 2025

franconi requested review from pfps, niklasl, afs, pchampin, hartig, Antoine-Zimmermann, doerthe and rat10 January 27, 2025 18:16

franconi added the spec:substantive Change in the spec affecting its normative content (class 3) –see also spec:bug, spec:new-feature label Jan 27, 2025

TallTed suggested changes Jan 27, 2025

View reviewed changes

spec/index.html Outdated Show resolved Hide resolved

franconi mentioned this pull request Jan 27, 2025

modification to new semantics to permit entailment that every triple term is a proposition #69

Closed

gkellogg reviewed Jan 27, 2025

View reviewed changes

spec/index.html Outdated Show resolved Hide resolved

franconi added 2 commits January 28, 2025 10:30

Fixed HTML + RTD -> RDF

4204f21

Fixed more HTML

ad4faa5

Minor fix in text for better English

3999bd8

Co-authored-by: Ted Thibodeau Jr <[email protected]>

TallTed reviewed Jan 28, 2025

View reviewed changes

spec/index.html Outdated Show resolved Hide resolved

franconi mentioned this pull request Jan 29, 2025

Fixing the definition of instance of a graph to consider triple terms #71

Closed

pchampin mentioned this pull request Jan 29, 2025

"RDF term" definition should be exported w3c/rdf-concepts#152

Open

franconi mentioned this pull request Jan 29, 2025

New entailment pattern rdfD1 #70

Open

pfps reviewed Jan 29, 2025

View reviewed changes

spec/index.html Outdated Show resolved Hide resolved

pfps requested changes Jan 29, 2025

View reviewed changes

spec/index.html Show resolved Hide resolved

pchampin approved these changes Jan 29, 2025

View reviewed changes

franconi added 2 commits January 29, 2025 16:07

Better English: added comma

4e1cef6

Added axiomatic triple in RDF for rdf:reifies

8487a13

Moved the "appears in" definition before its first use.

c5914b4

niklasl reviewed Jan 29, 2025

View reviewed changes

spec/index.html Outdated Show resolved Hide resolved

franconi and others added 3 commits January 29, 2025 16:23

"RDF term" definition is exported

fd09c1f

Co-authored-by: Pierre-Antoine Champin <[email protected]>

Added the reference to RTDF symmetric triple

d734184

Changed rdf:range to rdfs:range

b9c0e07

Co-authored-by: Niklas Lindström <[email protected]>

pfps approved these changes Jan 29, 2025

View reviewed changes

pchampin reviewed Jan 29, 2025

View reviewed changes

TallTed reviewed Jan 29, 2025

View reviewed changes

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Update RDF-semantics with the liberal baseline RDF and RDFS semantics #74

Update RDF-semantics with the liberal baseline RDF and RDFS semantics #74

franconi commented Jan 27, 2025 •

edited by pr-preview bot

Loading

TallTed left a comment

gkellogg commented Jan 27, 2025

franconi commented Jan 28, 2025

pfps left a comment

pchampin left a comment

pchampin Jan 29, 2025

pchampin Jan 29, 2025

franconi Jan 29, 2025

pchampin Jan 29, 2025

franconi Jan 29, 2025

pchampin Jan 29, 2025

franconi Jan 29, 2025

pchampin Jan 29, 2025

pchampin Jan 29, 2025

franconi Jan 29, 2025

rat10 commented Jan 29, 2025

franconi commented Jan 29, 2025

pfps commented Jan 29, 2025 •

edited

Loading

pfps left a comment

franconi commented Jan 29, 2025

pchampin Jan 29, 2025

TallTed Jan 29, 2025

TallTed Jan 29, 2025

TallTed Jan 29, 2025

	<th><strong>if the triple or triple term <a>appears in</a> S</strong></th>
	<th><strong>if the triple <a>appears in</a> S</strong></th>

	<th >If the triple or triple term <a>appears in</a> S:</th>
	<th >If the triple <a>appears in</a> S:</th>

	<td class="othertable">the graph S with the appearance replaced with <br/>
	<td class="othertable">

		exist x,y,z such that RE(x,z,y)=r, or exists x such that < x,r > is in IEXT([I+A](<code>rdf:reifies</code>)), <br/> then < r,[I+A](<code>rdfs:Proposition</code>)> is in IEXT([I+A](<code>rdf:type</code>))</p></td>
		</tr>

		An RDF term <dfn>appears in</dfn> a graph G if it is in the set TM(G), defined inductively as follows: (1) for each triple in the graph G, its subject, predicate, and object are in TM(G); (2) for each triple term in TH(G), its subject, predicate, and object are in TM(G).<br/>
		A triple <b><i>appears in</i></b> a graph G if it is in G or if the triple term composed by its subject, predicate, and object appears in G.

Update RDF-semantics with the liberal baseline RDF and RDFS semantics #74

Are you sure you want to change the base?

Update RDF-semantics with the liberal baseline RDF and RDFS semantics #74

Conversation

franconi commented Jan 27, 2025 • edited by pr-preview bot Loading

TallTed left a comment

Choose a reason for hiding this comment

gkellogg commented Jan 27, 2025

franconi commented Jan 28, 2025

pfps left a comment

Choose a reason for hiding this comment

pchampin 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

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

rat10 commented Jan 29, 2025

franconi commented Jan 29, 2025

pfps commented Jan 29, 2025 • edited Loading

pfps left a comment

Choose a reason for hiding this comment

franconi commented Jan 29, 2025

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

franconi commented Jan 27, 2025 •

edited by pr-preview bot

Loading

pfps commented Jan 29, 2025 •

edited

Loading