Rhetorical annotation and LOD

classic Classic list List threaded Threaded
17 messages Options
Reply | Threaded
Open this post in threaded view
|

Rhetorical annotation and LOD

Paolo Monella
Dear all,

I'm tutoring a student for an MA thesis. She's creating TEI XML
rhetorical annotation on a Latin text, like this:

[in the teiHeader:]
                  <interp xml:id="anafora"
sameAs="https://www.wikidata.org/wiki/Q486050">Anafora</interp>

[in the body:]
         <lb n="123.17"/>'scripsit causas: quia, <seg ana="#anafora">cum
semel appellassem, cum <supplied>iterum appellassem,
         <lb n="123.18"/>cum</supplied> tertio appellassem</seg>, non
corrupi'. [etc.]

We're trying to also express these statemtents (passage xyx has
rhetorical figure: anaphora) in LOD. Do you have any suggestions?

I did some research, but I can't find hasRhetoricalFigure anywhere except in
http://www.repositori.com/sw/onto/RhetoricalDevices.owl

I'm new to LOD, and I'm afraid that the following code doesn't even make
sense:

<?xml version="1.0"?>
<rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
   <rdf:Description rdf:about="annotated-document.xml#id-of-annotated-span">
     <rdf:instanceOf>
     https://www.wikidata.org/wiki/Q486050
     </rdf:instanceOf>
   </rdf:Description>
</rdf:RDF>

Thank you,
Paolo
Reply | Threaded
Open this post in threaded view
|

Re: Rhetorical annotation and LOD

Elisa Beshero-Bondar-2
Dear Paolo,
I’m new to LOD-ifying TEI, too, and I’m looking forward to reading the wisdom of the list, but I think I see what isn’t making sense in the RDF expression. As I understand it, when generating LOD expressions from TEI, we are explicitly spelling out relationships that we express in abbreviated ways.  Here, every <seg> element has an @ana whose value points to a wikidata link identified up in the teiHeader. Where you have:
<rdf:Description rdf:about="annotated-document.xml#id-of-annotated-span">
I think you want the value of @rdf:about to be a full pointer to the seg element, which would mean you would need an @xml:id on each seg to be able to isolate it. I think you will want to find a good system to set an @xml:id on your seg: the rule is an xml:id must contain some text and may contain numbers, and you might want to automatically generate these (with an XSLT identity transformation or an XQuery typeswitch) based on their position or location in the document, or something that makes sense to you to distinguish each one. 

You decode its relationship just as you are doing: follow the @ana back up to its definition and association with the wikidata link.

Once you figure out the relationship expression, you want to write XSLT or XQuery to output the RDF file, which will be kind of a long catalog of the <seg> ids and their associations with authoritative rhetorical definitions. 

Hope this helps!
Elisa
--
Elisa Beshero-Bondar, PhD 
Director, Center for the Digital Text
Associate Professor of English 
University of Pittsburgh at Greensburg
150 Finoli Drive, Greensburg, PA 15601 USA
E-mail: [hidden email] | Development site: http://newtfire.org

Typeset by hand on my iPad

On Feb 11, 2018, at 4:23 AM, Paolo Monella <[hidden email]> wrote:

Dear all,

I'm tutoring a student for an MA thesis. She's creating TEI XML rhetorical annotation on a Latin text, like this:

[in the teiHeader:]
                <interp xml:id="anafora" sameAs="https://www.wikidata.org/wiki/Q486050">Anafora</interp>

[in the body:]
       <lb n="123.17"/>'scripsit causas: quia, <seg ana="#anafora">cum semel appellassem, cum <supplied>iterum appellassem,
       <lb n="123.18"/>cum</supplied> tertio appellassem</seg>, non corrupi'. [etc.]

We're trying to also express these statemtents (passage xyx has rhetorical figure: anaphora) in LOD. Do you have any suggestions?

I did some research, but I can't find hasRhetoricalFigure anywhere except in
http://www.repositori.com/sw/onto/RhetoricalDevices.owl

I'm new to LOD, and I'm afraid that the following code doesn't even make sense:

<?xml version="1.0"?>
<rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
 <rdf:Description rdf:about="annotated-document.xml#id-of-annotated-span">
   <rdf:instanceOf>
   https://www.wikidata.org/wiki/Q486050
   </rdf:instanceOf>
 </rdf:Description>
</rdf:RDF>

Thank you,
Paolo
Reply | Threaded
Open this post in threaded view
|

Re: Rhetorical annotation and LOD

Lou Burnard-6
In reply to this post by Paolo Monella

I'm always happy to expose myself to ridicule on this list, so I would like to know why the following wouldn't be easier and just as explicit:


   <lb n="123.17"/>'scripsit causas: quia, <seg ana="https://www.wikidata.org/wiki/Q486050">cum semel appellassem, cum <supplied>iterum appellassem,
        <lb n="123.18"/>cum</supplied> tertio appellassem</seg>, non corrupi'. [etc.]

Or, of course, you could do a prefixdef for "https://wikidata.org/wiki" and simply say

<seg ana="wikidata:Q486050">


L




On 11/02/18 09:23, Paolo Monella wrote:
Dear all,

I'm tutoring a student for an MA thesis. She's creating TEI XML rhetorical annotation on a Latin text, like this:

[in the teiHeader:]
                 <interp xml:id="anafora" sameAs="https://www.wikidata.org/wiki/Q486050">Anafora</interp>

[in the body:]
        <lb n="123.17"/>'scripsit causas: quia, <seg ana="#anafora">cum semel appellassem, cum <supplied>iterum appellassem,
        <lb n="123.18"/>cum</supplied> tertio appellassem</seg>, non corrupi'. [etc.]

We're trying to also express these statemtents (passage xyx has rhetorical figure: anaphora) in LOD. Do you have any suggestions?

I did some research, but I can't find hasRhetoricalFigure anywhere except in
http://www.repositori.com/sw/onto/RhetoricalDevices.owl

I'm new to LOD, and I'm afraid that the following code doesn't even make sense:

<?xml version="1.0"?>
<rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
  <rdf:Description rdf:about="annotated-document.xml#id-of-annotated-span">
    <rdf:instanceOf>
    https://www.wikidata.org/wiki/Q486050
    </rdf:instanceOf>
  </rdf:Description>
</rdf:RDF>

Thank you,
Paolo

Reply | Threaded
Open this post in threaded view
|

Re: Rhetorical annotation and LOD

Elisa Beshero-Bondar-2
In reply to this post by Elisa Beshero-Bondar-2
PS: I should have written:
Once you figure out the relationship expression, you want to write XSLT or XQuery to output the RDF file, which will be kind of a long catalog of full pointers to the <seg> elements (a URL with a pointer to the location of each seg) and their associations with authoritative rhetorical definitions. 
Unpacking the abbreviations of ids and anas is, as I understand it, important to generating literal RDF expressions to help bring data to the “surface” of your project on the linked web of data. 

Elisa

--
Elisa Beshero-Bondar, PhD 
Director, Center for the Digital Text
Associate Professor of English 
University of Pittsburgh at Greensburg
150 Finoli Drive, Greensburg, PA 15601 USA
E-mail: [hidden email] | Development site: http://newtfire.org

Typeset by hand on my iPad

On Feb 11, 2018, at 10:13 AM, Elisa <[hidden email]> wrote:

Dear Paolo,
I’m new to LOD-ifying TEI, too, and I’m looking forward to reading the wisdom of the list, but I think I see what isn’t making sense in the RDF expression. As I understand it, when generating LOD expressions from TEI, we are explicitly spelling out relationships that we express in abbreviated ways.  Here, every <seg> element has an @ana whose value points to a wikidata link identified up in the teiHeader. Where you have:
<rdf:Description rdf:about="annotated-document.xml#id-of-annotated-span">
I think you want the value of @rdf:about to be a full pointer to the seg element, which would mean you would need an @xml:id on each seg to be able to isolate it. I think you will want to find a good system to set an @xml:id on your seg: the rule is an xml:id must contain some text and may contain numbers, and you might want to automatically generate these (with an XSLT identity transformation or an XQuery typeswitch) based on their position or location in the document, or something that makes sense to you to distinguish each one. 

You decode its relationship just as you are doing: follow the @ana back up to its definition and association with the wikidata link.

Once you figure out the relationship expression, you want to write XSLT or XQuery to output the RDF file, which will be kind of a long catalog of the <seg> ids and their associations with authoritative rhetorical definitions. 

Hope this helps!
Elisa
--
Elisa Beshero-Bondar, PhD 
Director, Center for the Digital Text
Associate Professor of English 
University of Pittsburgh at Greensburg
150 Finoli Drive, Greensburg, PA 15601 USA
E-mail: [hidden email] | Development site: http://newtfire.org

Typeset by hand on my iPad

On Feb 11, 2018, at 4:23 AM, Paolo Monella <[hidden email]> wrote:

Dear all,

I'm tutoring a student for an MA thesis. She's creating TEI XML rhetorical annotation on a Latin text, like this:

[in the teiHeader:]
                <interp xml:id="anafora" sameAs="https://www.wikidata.org/wiki/Q486050">Anafora</interp>

[in the body:]
       <lb n="123.17"/>'scripsit causas: quia, <seg ana="#anafora">cum semel appellassem, cum <supplied>iterum appellassem,
       <lb n="123.18"/>cum</supplied> tertio appellassem</seg>, non corrupi'. [etc.]

We're trying to also express these statemtents (passage xyx has rhetorical figure: anaphora) in LOD. Do you have any suggestions?

I did some research, but I can't find hasRhetoricalFigure anywhere except in
http://www.repositori.com/sw/onto/RhetoricalDevices.owl

I'm new to LOD, and I'm afraid that the following code doesn't even make sense:

<?xml version="1.0"?>
<rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
 <rdf:Description rdf:about="annotated-document.xml#id-of-annotated-span">
   <rdf:instanceOf>
   https://www.wikidata.org/wiki/Q486050
   </rdf:instanceOf>
 </rdf:Description>
</rdf:RDF>

Thank you,
Paolo
Reply | Threaded
Open this post in threaded view
|

Re: Rhetorical annotation and LOD

Elisa Beshero-Bondar
In reply to this post by Lou Burnard-6
Hi Lou-- That raises the question of whether the project considers the wikidata definitions to be permanently authoritative. If you've got those pointers down in every seg, and the wiki goes away or changes its numbers, well it's harder to rip them out later. If you've got the relationships simply defined up the teiHeader, it's less code to have to change if the wikidata goes away someday and the pointers need to resolve differently. 

I'm not just raising that to be contrarian--That's a serious issue in the LOD world! (Nought in this age is lasting...Nought endures but mutability...) The TEI encoding of #anaphora on Paolo's project might be best left on its own, and the rhetorical definitions left relatively easy to redefine should the project developers decide they want to change the links out or point to a better definitive resource. 

Elisa



On Sun, Feb 11, 2018 at 10:20 AM, Lou Burnard <[hidden email]> wrote:

I'm always happy to expose myself to ridicule on this list, so I would like to know why the following wouldn't be easier and just as explicit:


   <lb n="123.17"/>'scripsit causas: quia, <seg ana="https://www.wikidata.org/wiki/Q486050">cum semel appellassem, cum <supplied>iterum appellassem,
        <lb n="123.18"/>cum</supplied> tertio appellassem</seg>, non corrupi'. [etc.]

Or, of course, you could do a prefixdef for "https://wikidata.org/wiki" and simply say

<seg ana="wikidata:Q486050">


L




On 11/02/18 09:23, Paolo Monella wrote:
Dear all,

I'm tutoring a student for an MA thesis. She's creating TEI XML rhetorical annotation on a Latin text, like this:

[in the teiHeader:]
                 <interp xml:id="anafora" sameAs="https://www.wikidata.org/wiki/Q486050">Anafora</interp>

[in the body:]
        <lb n="123.17"/>'scripsit causas: quia, <seg ana="#anafora">cum semel appellassem, cum <supplied>iterum appellassem,
        <lb n="123.18"/>cum</supplied> tertio appellassem</seg>, non corrupi'. [etc.]

We're trying to also express these statemtents (passage xyx has rhetorical figure: anaphora) in LOD. Do you have any suggestions?

I did some research, but I can't find hasRhetoricalFigure anywhere except in
http://www.repositori.com/sw/onto/RhetoricalDevices.owl

I'm new to LOD, and I'm afraid that the following code doesn't even make sense:

<?xml version="1.0"?>
<rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
  <rdf:Description rdf:about="annotated-document.xml#id-of-annotated-span">
    <rdf:instanceOf>
    https://www.wikidata.org/wiki/Q486050
    </rdf:instanceOf>
  </rdf:Description>
</rdf:RDF>

Thank you,
Paolo




--
Elisa Beshero-Bondar, PhD
Director, Center for the Digital Text | Associate Professor of English
University of Pittsburgh at Greensburg | Humanities Division
150 Finoli Drive
Greensburg, PA  15601  USA
E-mail:[hidden email]
Development site: http://newtfire.org
Reply | Threaded
Open this post in threaded view
|

Re: Rhetorical annotation and LOD

Lou Burnard-6

Good point, but it doesn't convince me.

This project says "I want to call those things ANAPHORA (oh, and by ANAPHORA I mean the same thing Wikidata calls Q486050)"

So if the meaning of "Wikidata Q486050" suddenly ceases to exist for some reason, all I know is that this project considers these things to be the same sort of thing (ANAPHORA) but I don't know what that means. In which case, I might as well have called them "FRIEDTOMATOES" or even "wikidata:W486050": I wouldn't be any better or worse off.


On 11/02/18 15:48, Elisa Beshero-Bondar wrote:
Hi Lou-- That raises the question of whether the project considers the
wikidata definitions to be permanently authoritative. If you've got those
pointers down in every seg, and the wiki goes away or changes its numbers,
well it's harder to rip them out later. If you've got the relationships
simply defined up the teiHeader, it's less code to have to change if the
wikidata goes away someday and the pointers need to resolve differently.

I'm not just raising that to be contrarian--That's a serious issue in the
LOD world! (Nought in this age is lasting...Nought endures but
mutability...) The TEI encoding of #anaphora on Paolo's project might be
best left on its own, and the rhetorical definitions left relatively easy
to redefine should the project developers decide they want to change the
links out or point to a better definitive resource.

Elisa



On Sun, Feb 11, 2018 at 10:20 AM, Lou Burnard [hidden email]
wrote:

I'm always happy to expose myself to ridicule on this list, so I would
like to know why the following wouldn't be easier and just as explicit:


   <lb n="123.17"/>'scripsit causas: quia, <seg ana="
https://www.wikidata.org/wiki/Q486050">cum semel appellassem, cum
<supplied>iterum appellassem,
        <lb n="123.18"/>cum</supplied> tertio appellassem</seg>, non
corrupi'. [etc.]

Or, of course, you could do a prefixdef for "https://wikidata.org/wiki"
<https://wikidata.org/wiki> and simply say

<seg ana="wikidata:Q486050">


L




On 11/02/18 09:23, Paolo Monella wrote:

Dear all,

I'm tutoring a student for an MA thesis. She's creating TEI XML rhetorical
annotation on a Latin text, like this:

[in the teiHeader:]
                 <interp xml:id="anafora" sameAs="https://www.wikidata.
org/wiki/Q486050" <https://www.wikidata.org/wiki/Q486050>>Anafora</interp>


[in the body:]
        <lb n="123.17"/>'scripsit causas: quia, <seg ana="#anafora">cum
semel appellassem, cum <supplied>iterum appellassem,
        <lb n="123.18"/>cum</supplied> tertio appellassem</seg>, non
corrupi'. [etc.]

We're trying to also express these statemtents (passage xyx has rhetorical
figure: anaphora) in LOD. Do you have any suggestions?

I did some research, but I can't find hasRhetoricalFigure anywhere except
in
http://www.repositori.com/sw/onto/RhetoricalDevices.owl

I'm new to LOD, and I'm afraid that the following code doesn't even make
sense:

<?xml version="1.0"?>
<rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
<http://www.w3.org/1999/02/22-rdf-syntax-ns#>
  <rdf:Description rdf:about="annotated-document.xml#id-of-annotated-span">

    <rdf:instanceOf>
    https://www.wikidata.org/wiki/Q486050
    </rdf:instanceOf>
  </rdf:Description>
</rdf:RDF>

Thank you,
Paolo





Reply | Threaded
Open this post in threaded view
|

Re: Rhetorical annotation and LOD

Elisa Beshero-Bondar
And I am not "convinced" by your counterpoint, Lou. You seem to be qualifying the project team's usage of "#anaphora" as somehow just as meaningful as any random string of text. It isn't--its meaning, as Paolo indicated, is defined in the teiHeader where such things are quite properly defined. 

That is, there isn't anything wrong with their current practice, and quite a lot that's right about it. I'll even add something more: What if the project developers wanted to add linkages to other rhetorical databanks to their project, to encode multiple definitions of anaphora? Would you then argue that each of those should literally appear in the @ana on each and every seg? 

Elisa

On Sun, Feb 11, 2018 at 10:57 AM, Lou Burnard <[hidden email]> wrote:

Good point, but it doesn't convince me.

This project says "I want to call those things ANAPHORA (oh, and by ANAPHORA I mean the same thing Wikidata calls Q486050)"

So if the meaning of "Wikidata Q486050" suddenly ceases to exist for some reason, all I know is that this project considers these things to be the same sort of thing (ANAPHORA) but I don't know what that means. In which case, I might as well have called them "FRIEDTOMATOES" or even "wikidata:W486050": I wouldn't be any better or worse off.


On 11/02/18 15:48, Elisa Beshero-Bondar wrote:
Hi Lou-- That raises the question of whether the project considers the
wikidata definitions to be permanently authoritative. If you've got those
pointers down in every seg, and the wiki goes away or changes its numbers,
well it's harder to rip them out later. If you've got the relationships
simply defined up the teiHeader, it's less code to have to change if the
wikidata goes away someday and the pointers need to resolve differently.

I'm not just raising that to be contrarian--That's a serious issue in the
LOD world! (Nought in this age is lasting...Nought endures but
mutability...) The TEI encoding of #anaphora on Paolo's project might be
best left on its own, and the rhetorical definitions left relatively easy
to redefine should the project developers decide they want to change the
links out or point to a better definitive resource.

Elisa



On Sun, Feb 11, 2018 at 10:20 AM, Lou Burnard [hidden email]
wrote:

I'm always happy to expose myself to ridicule on this list, so I would like to know why the following wouldn't be easier and just as explicit: <lb n="123.17"/>'scripsit causas: quia, <seg ana=" https://www.wikidata.org/wiki/Q486050">cum semel appellassem, cum <supplied>iterum appellassem, <lb n="123.18"/>cum</supplied> tertio appellassem</seg>, non corrupi'. [etc.] Or, of course, you could do a prefixdef for "https://wikidata.org/wiki" <https://wikidata.org/wiki> and simply say <seg ana="wikidata:Q486050"> L On 11/02/18 09:23, Paolo Monella wrote: Dear all, I'm tutoring a student for an MA thesis. She's creating TEI XML rhetorical annotation on a Latin text, like this: [in the teiHeader:] <interp xml:id="anafora" sameAs=
"https://www.wikidata. org/wiki/Q486050" <https://www.wikidata.org/wiki/Q486050>>Anafora</interp> [in the body:] <lb n="123.17"/>'scripsit causas: quia, <seg ana="#anafora">cum semel appellassem, cum <supplied>iterum appellassem, <lb n="123.18"/>cum</supplied> tertio appellassem</seg>, non corrupi'. [etc.] We're trying to also express these statemtents (passage xyx has rhetorical figure: anaphora) in LOD. Do you have any suggestions? I did some research, but I can't find hasRhetoricalFigure anywhere except in http://www.repositori.com/sw/onto/RhetoricalDevices.owl I'm new to LOD, and I'm afraid that the following code doesn't even make sense: <?xml version="1.0"?> <rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" <http://www.w3.org/1999/02/22-rdf-syntax-ns#> <rdf:Description rdf:about="annotated-document.xml#id-of-annotated-span"> <rdf:instanceOf> https://www.wikidata.org/wiki/Q486050 </rdf:instanceOf> </rdf:Description> </rdf:RDF> Thank you, Paolo

    




--
Elisa Beshero-Bondar, PhD
Director, Center for the Digital Text | Associate Professor of English
University of Pittsburgh at Greensburg | Humanities Division
150 Finoli Drive
Greensburg, PA  15601  USA
E-mail:[hidden email]
Development site: http://newtfire.org
Reply | Threaded
Open this post in threaded view
|

Re: Rhetorical annotation and LOD

Lou Burnard-6

The argument that one might want to associate ANAPHORA with more than one definition is certainly more persuasive, though I don't know that such is the case here.

I wasn't making any criticism of this particular project, by the way; merely musing aloud about simpler ways of expressing a very simple idea.


On 11/02/18 16:03, Elisa Beshero-Bondar wrote:
And I am not "convinced" by your counterpoint, Lou. You seem to be
qualifying the project team's usage of "#anaphora" as somehow just as
meaningful as any random string of text. It isn't--its meaning, as Paolo
indicated, is defined in the teiHeader where such things are quite properly
defined.

That is, there isn't anything wrong with their current practice, and quite
a lot that's right about it. I'll even add something more: What if the
project developers wanted to add linkages to other rhetorical databanks to
their project, to encode multiple definitions of anaphora? Would you then
argue that each of those should literally appear in the @ana on each and
every seg?

Elisa

On Sun, Feb 11, 2018 at 10:57 AM, Lou Burnard [hidden email]
wrote:

Good point, but it doesn't convince me.

This project says "I want to call those things ANAPHORA (oh, and by
ANAPHORA I mean the same thing Wikidata calls Q486050)"

So if the meaning of "Wikidata Q486050" suddenly ceases to exist for some
reason, all I know is that this project considers these things to be the
same sort of thing (ANAPHORA) but I don't know what that means. In which
case, I might as well have called them "FRIEDTOMATOES" or even
"wikidata:W486050": I wouldn't be any better or worse off.


On 11/02/18 15:48, Elisa Beshero-Bondar wrote:

Hi Lou-- That raises the question of whether the project considers the
wikidata definitions to be permanently authoritative. If you've got those
pointers down in every seg, and the wiki goes away or changes its numbers,
well it's harder to rip them out later. If you've got the relationships
simply defined up the teiHeader, it's less code to have to change if the
wikidata goes away someday and the pointers need to resolve differently.

I'm not just raising that to be contrarian--That's a serious issue in the
LOD world! (Nought in this age is lasting...Nought endures but
mutability...) The TEI encoding of #anaphora on Paolo's project might be
best left on its own, and the rhetorical definitions left relatively easy
to redefine should the project developers decide they want to change the
links out or point to a better definitive resource.

Elisa



On Sun, Feb 11, 2018 at 10:20 AM, Lou Burnard [hidden email] [hidden email]
wrote:


I'm always happy to expose myself to ridicule on this list, so I would
like to know why the following wouldn't be easier and just as explicit:


   <lb n="123.17"/>'scripsit causas: quia, <seg ana="
https://www.wikidata.org/wiki/Q486050" <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.wikidata.org%2Fwiki%2FQ486050&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=k%2B4vRD%2FXWq9beqs4yqxdmJ1SoX1NjNZzRA%2BWG8QO9IA%3D&reserved=0>>cum semel appellassem, cum
<supplied>iterum appellassem,
        <lb n="123.18"/>cum</supplied> tertio appellassem</seg>, non
corrupi'. [etc.]

Or, of course, you could do a prefixdef for "https://wikidata.org/wiki" <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwikidata.org%2Fwiki&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=dRVPTdRfLncAJBXgSdjBR%2BpE0WgXVK0FBt2APHyhojE%3D&reserved=0><https://wikidata.org/wiki> <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwikidata.org%2Fwiki&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=dRVPTdRfLncAJBXgSdjBR%2BpE0WgXVK0FBt2APHyhojE%3D&reserved=0> and simply say

<seg ana="wikidata:Q486050">


L




On 11/02/18 09:23, Paolo Monella wrote:

Dear all,

I'm tutoring a student for an MA thesis. She's creating TEI XML rhetorical
annotation on a Latin text, like this:

[in the teiHeader:]
                 <interp xml:id="anafora" sameAs=
"https://www.wikidata.
org/wiki/Q486050" <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.wikidata.org%2Fwiki%2FQ486050&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=k%2B4vRD%2FXWq9beqs4yqxdmJ1SoX1NjNZzRA%2BWG8QO9IA%3D&reserved=0> <https://www.wikidata.org/wiki/Q486050> <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.wikidata.org%2Fwiki%2FQ486050&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=k%2B4vRD%2FXWq9beqs4yqxdmJ1SoX1NjNZzRA%2BWG8QO9IA%3D&reserved=0>>Anafora</interp>


[in the body:]
        <lb n="123.17"/>'scripsit causas: quia, <seg ana="#anafora">cum
semel appellassem, cum <supplied>iterum appellassem,
        <lb n="123.18"/>cum</supplied> tertio appellassem</seg>, non
corrupi'. [etc.]

We're trying to also express these statemtents (passage xyx has rhetorical
figure: anaphora) in LOD. Do you have any suggestions?

I did some research, but I can't find hasRhetoricalFigure anywhere except
inhttp://www.repositori.com/sw/onto/RhetoricalDevices.owl <https://na01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.repositori.com%2Fsw%2Fonto%2FRhetoricalDevices.owl&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=OnrpazEDej4bqF8sGyF0u1kQxOJMpzeyFFyqdzBaykg%3D&reserved=0>

I'm new to LOD, and I'm afraid that the following code doesn't even make
sense:

<?xml version="1.0"?>
<rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" <https://na01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.w3.org%2F1999%2F02%2F22-rdf-syntax-ns%23&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=G7CdbsEK%2BgeLk3mZpAegalg5cYKEJJB0tWDmAYorz5E%3D&reserved=0><http://www.w3.org/1999/02/22-rdf-syntax-ns#> <https://na01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.w3.org%2F1999%2F02%2F22-rdf-syntax-ns%23&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=G7CdbsEK%2BgeLk3mZpAegalg5cYKEJJB0tWDmAYorz5E%3D&reserved=0>
  <rdf:Description rdf:about="annotated-document.xml#id-of-annotated-span">

    <rdf:instanceOf>
    https://www.wikidata.org/wiki/Q486050 <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.wikidata.org%2Fwiki%2FQ486050&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=k%2B4vRD%2FXWq9beqs4yqxdmJ1SoX1NjNZzRA%2BWG8QO9IA%3D&reserved=0>
    </rdf:instanceOf>
  </rdf:Description>
</rdf:RDF>

Thank you,
Paolo








Reply | Threaded
Open this post in threaded view
|

Re: Rhetorical annotation and LOD

Elisa Beshero-Bondar
Lou-- All in good "musing"--which is one thing we do well on this list! :-)  One thing I think our little debate is demonstrating is a larger tension between TEI and LOD. Each of these considers itself a system to encode information sustainably and interchangeably, and each is a little at odds with the other. TEI is not structured like RDF, but the relationships we encode can be transformed and expressed in those terms. Some have argued that TEI is "already" LOD, but practitioners of LOD complain that the values of pointers in our attributes need to be fully spelled out in each literal location for that to be achieved (on each and every seg element that has to be decoded to find its linked-open-data relationship). 

We can do that, but the LOD community isn't as stable as some of us would like it to be, and perhaps the expression and cataloging of differences is as lasting as we would hope of our linked networks of information. Some of us in the TEI world ind LOD a little uncertain and vexingly limited (as Paolo indicated--there's just one available vocabulary to work with, and maybe we're hoping it will be improved someday). We may want to participate in LOD following its vocabularies, but it might be that those will change, or more and better will become available. I don't like the sense that inventors of (not-so?) controlled vocabularies project a lasting stability to their structures, and so I guess I favor encodings that are a little more open to "mutability" in the LOD world. 

Anyway, for playing in a friendly way with LOD, and maintaining a sense of internal consistency within a project, I think the teiHeader is a good place to define relationships that might have to change later.

Cheers,

Elisa


On Sun, Feb 11, 2018 at 11:10 AM, Lou Burnard <[hidden email]> wrote:

The argument that one might want to associate ANAPHORA with more than one definition is certainly more persuasive, though I don't know that such is the case here.

I wasn't making any criticism of this particular project, by the way; merely musing aloud about simpler ways of expressing a very simple idea.


On 11/02/18 16:03, Elisa Beshero-Bondar wrote:
And I am not "convinced" by your counterpoint, Lou. You seem to be
qualifying the project team's usage of "#anaphora" as somehow just as
meaningful as any random string of text. It isn't--its meaning, as Paolo
indicated, is defined in the teiHeader where such things are quite properly
defined.

That is, there isn't anything wrong with their current practice, and quite
a lot that's right about it. I'll even add something more: What if the
project developers wanted to add linkages to other rhetorical databanks to
their project, to encode multiple definitions of anaphora? Would you then
argue that each of those should literally appear in the @ana on each and
every seg?

Elisa

On Sun, Feb 11, 2018 at 10:57 AM, Lou Burnard [hidden email]
wrote:

Good point, but it doesn't convince me.

This project says "I want to call those things ANAPHORA (oh, and by
ANAPHORA I mean the same thing Wikidata calls Q486050)"

So if the meaning of "Wikidata Q486050" suddenly ceases to exist for some
reason, all I know is that this project considers these things to be the
same sort of thing (ANAPHORA) but I don't know what that means. In which
case, I might as well have called them "FRIEDTOMATOES" or even
"wikidata:W486050": I wouldn't be any better or worse off.


On 11/02/18 15:48, Elisa Beshero-Bondar wrote:

Hi Lou-- That raises the question of whether the project considers the
wikidata definitions to be permanently authoritative. If you've got those
pointers down in every seg, and the wiki goes away or changes its numbers,
well it's harder to rip them out later. If you've got the relationships
simply defined up the teiHeader, it's less code to have to change if the
wikidata goes away someday and the pointers need to resolve differently.

I'm not just raising that to be contrarian--That's a serious issue in the
LOD world! (Nought in this age is lasting...Nought endures but
mutability...) The TEI encoding of #anaphora on Paolo's project might be
best left on its own, and the rhetorical definitions left relatively easy
to redefine should the project developers decide they want to change the
links out or point to a better definitive resource.

Elisa



On Sun, Feb 11, 2018 at 10:20 AM, Lou Burnard [hidden email] [hidden email]
wrote:


I'm always happy to expose myself to ridicule on this list, so I would
like to know why the following wouldn't be easier and just as explicit:


   <lb n="123.17"/>'scripsit causas: quia, <seg ana="
https://www.wikidata.org/wiki/Q486050" <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.wikidata.org%2Fwiki%2FQ486050&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=k%2B4vRD%2FXWq9beqs4yqxdmJ1SoX1NjNZzRA%2BWG8QO9IA%3D&reserved=0>>cum semel appellassem, cum
<supplied>iterum appellassem,
        <lb n="123.18"/>cum</supplied> tertio appellassem</seg>, non
corrupi'. [etc.]

Or, of course, you could do a prefixdef for "https://wikidata.org/wiki" <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwikidata.org%2Fwiki&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=dRVPTdRfLncAJBXgSdjBR%2BpE0WgXVK0FBt2APHyhojE%3D&reserved=0><https://wikidata.org/wiki> <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwikidata.org%2Fwiki&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=dRVPTdRfLncAJBXgSdjBR%2BpE0WgXVK0FBt2APHyhojE%3D&reserved=0> and simply say

<seg ana="wikidata:Q486050">


L




On 11/02/18 09:23, Paolo Monella wrote:

Dear all,

I'm tutoring a student for an MA thesis. She's creating TEI XML rhetorical
annotation on a Latin text, like this:

[in the teiHeader:]
                 <interp xml:id="anafora" sameAs=
"https://www.wikidata.
org/wiki/Q486050" <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.wikidata.org%2Fwiki%2FQ486050&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=k%2B4vRD%2FXWq9beqs4yqxdmJ1SoX1NjNZzRA%2BWG8QO9IA%3D&reserved=0> <https://www.wikidata.org/wiki/Q486050> <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.wikidata.org%2Fwiki%2FQ486050&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=k%2B4vRD%2FXWq9beqs4yqxdmJ1SoX1NjNZzRA%2BWG8QO9IA%3D&reserved=0>>Anafora</interp>


[in the body:]
        <lb n="123.17"/>'scripsit causas: quia, <seg ana="#anafora">cum
semel appellassem, cum <supplied>iterum appellassem,
        <lb n="123.18"/>cum</supplied> tertio appellassem</seg>, non
corrupi'. [etc.]

We're trying to also express these statemtents (passage xyx has rhetorical
figure: anaphora) in LOD. Do you have any suggestions?

I did some research, but I can't find hasRhetoricalFigure anywhere except
inhttp://www.repositori.com/sw/onto/RhetoricalDevices.owl <https://na01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.repositori.com%2Fsw%2Fonto%2FRhetoricalDevices.owl&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=OnrpazEDej4bqF8sGyF0u1kQxOJMpzeyFFyqdzBaykg%3D&reserved=0>

I'm new to LOD, and I'm afraid that the following code doesn't even make
sense:

<?xml version="1.0"?>
<rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" <https://na01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.w3.org%2F1999%2F02%2F22-rdf-syntax-ns%23&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=G7CdbsEK%2BgeLk3mZpAegalg5cYKEJJB0tWDmAYorz5E%3D&reserved=0><http://www.w3.org/1999/02/22-rdf-syntax-ns#> <https://na01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.w3.org%2F1999%2F02%2F22-rdf-syntax-ns%23&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=G7CdbsEK%2BgeLk3mZpAegalg5cYKEJJB0tWDmAYorz5E%3D&reserved=0>
  <rdf:Description rdf:about="annotated-document.xml#id-of-annotated-span">

    <rdf:instanceOf>
    https://www.wikidata.org/wiki/Q486050 <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.wikidata.org%2Fwiki%2FQ486050&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=k%2B4vRD%2FXWq9beqs4yqxdmJ1SoX1NjNZzRA%2BWG8QO9IA%3D&reserved=0>
    </rdf:instanceOf>
  </rdf:Description>
</rdf:RDF>

Thank you,
Paolo







    




--
Elisa Beshero-Bondar, PhD
Director, Center for the Digital Text | Associate Professor of English
University of Pittsburgh at Greensburg | Humanities Division
150 Finoli Drive
Greensburg, PA  15601  USA
E-mail:[hidden email]
Development site: http://newtfire.org
Reply | Threaded
Open this post in threaded view
|

Re: Rhetorical annotation and LOD

richard light

On 11/02/2018 16:13, Elisa Beshero-Bondar wrote:
Lou-- All in good "musing"--which is one thing we do well on this list! :-)  One thing I think our little debate is demonstrating is a larger tension between TEI and LOD. Each of these considers itself a system to encode information sustainably and interchangeably, and each is a little at odds with the other. TEI is not structured like RDF, but the relationships we encode can be transformed and expressed in those terms. Some have argued that TEI is "already" LOD, but practitioners of LOD complain that the values of pointers in our attributes need to be fully spelled out in each literal location for that to be achieved (on each and every seg element that has to be decoded to find its linked-open-data relationship). 

We can do that, but the LOD community isn't as stable as some of us would like it to be, and perhaps the expression and cataloging of differences is as lasting as we would hope of our linked networks of information. Some of us in the TEI world ind LOD a little uncertain and vexingly limited (as Paolo indicated--there's just one available vocabulary to work with, and maybe we're hoping it will be improved someday). We may want to participate in LOD following its vocabularies, but it might be that those will change, or more and better will become available. I don't like the sense that inventors of (not-so?) controlled vocabularies project a lasting stability to their structures, and so I guess I favor encodings that are a little more open to "mutability" in the LOD world. 

Anyway, for playing in a friendly way with LOD, and maintaining a sense of internal consistency within a project, I think the teiHeader is a good place to define relationships that might have to change later.
I would agree with that approach.  I have it in mind that a generic XSLT which publishes the TEI Header as a set of RDF assertions could be a generally useful addition to our armoury.  The development of such a transform would be greatly aided by community agreement on where and how LOD URLs might be included within the Header.

While it is valid to be concerned about the stability of LOD URLs, I wold argue that one way to help ensure that they do remain stable is to make it publicly known that they are in active use.  It's something of a chicken and egg situation.  I too have reservations about RDF, but I don't see any alternative to LOD URLs as a means of achieving semantic interoperability across the TEI world and beyond it.

Something else to bear in mind is that such URLs can be more than an additional recording burden and a source of worry: they can also provide additional information 'for free'.  In the case we are discussing:
https://www.wikidata.org/wiki/Q486050
you don't get a lot beyond multiple translations of 'anaphora' (although that might be useful).  Something I wrote to facilitate the use of LD resources in an XSLT processing environment is a little 'forwarder' program, which simply adds a suitable Accept header to an HTTP request, so that you can use the document() function on the URL and get RDF/XML back.

Best wishes,

Richard


Cheers,

Elisa


On Sun, Feb 11, 2018 at 11:10 AM, Lou Burnard <[hidden email]> wrote:

The argument that one might want to associate ANAPHORA with more than one definition is certainly more persuasive, though I don't know that such is the case here.

I wasn't making any criticism of this particular project, by the way; merely musing aloud about simpler ways of expressing a very simple idea.


On 11/02/18 16:03, Elisa Beshero-Bondar wrote:
And I am not "convinced" by your counterpoint, Lou. You seem to be
qualifying the project team's usage of "#anaphora" as somehow just as
meaningful as any random string of text. It isn't--its meaning, as Paolo
indicated, is defined in the teiHeader where such things are quite properly
defined.

That is, there isn't anything wrong with their current practice, and quite
a lot that's right about it. I'll even add something more: What if the
project developers wanted to add linkages to other rhetorical databanks to
their project, to encode multiple definitions of anaphora? Would you then
argue that each of those should literally appear in the @ana on each and
every seg?

Elisa

On Sun, Feb 11, 2018 at 10:57 AM, Lou Burnard [hidden email]
wrote:

Good point, but it doesn't convince me.

This project says "I want to call those things ANAPHORA (oh, and by
ANAPHORA I mean the same thing Wikidata calls Q486050)"

So if the meaning of "Wikidata Q486050" suddenly ceases to exist for some
reason, all I know is that this project considers these things to be the
same sort of thing (ANAPHORA) but I don't know what that means. In which
case, I might as well have called them "FRIEDTOMATOES" or even
"wikidata:W486050": I wouldn't be any better or worse off.


On 11/02/18 15:48, Elisa Beshero-Bondar wrote:

Hi Lou-- That raises the question of whether the project considers the
wikidata definitions to be permanently authoritative. If you've got those
pointers down in every seg, and the wiki goes away or changes its numbers,
well it's harder to rip them out later. If you've got the relationships
simply defined up the teiHeader, it's less code to have to change if the
wikidata goes away someday and the pointers need to resolve differently.

I'm not just raising that to be contrarian--That's a serious issue in the
LOD world! (Nought in this age is lasting...Nought endures but
mutability...) The TEI encoding of #anaphora on Paolo's project might be
best left on its own, and the rhetorical definitions left relatively easy
to redefine should the project developers decide they want to change the
links out or point to a better definitive resource.

Elisa



On Sun, Feb 11, 2018 at 10:20 AM, Lou Burnard [hidden email] [hidden email]
wrote:


I'm always happy to expose myself to ridicule on this list, so I would
like to know why the following wouldn't be easier and just as explicit:


   <lb n="123.17"/>'scripsit causas: quia, <seg ana="
https://www.wikidata.org/wiki/Q486050" <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.wikidata.org%2Fwiki%2FQ486050&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=k%2B4vRD%2FXWq9beqs4yqxdmJ1SoX1NjNZzRA%2BWG8QO9IA%3D&reserved=0>>cum semel appellassem, cum
<supplied>iterum appellassem,
        <lb n="123.18"/>cum</supplied> tertio appellassem</seg>, non
corrupi'. [etc.]

Or, of course, you could do a prefixdef for "https://wikidata.org/wiki" <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwikidata.org%2Fwiki&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=dRVPTdRfLncAJBXgSdjBR%2BpE0WgXVK0FBt2APHyhojE%3D&reserved=0><https://wikidata.org/wiki> <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwikidata.org%2Fwiki&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=dRVPTdRfLncAJBXgSdjBR%2BpE0WgXVK0FBt2APHyhojE%3D&reserved=0> and simply say

<seg ana="wikidata:Q486050">


L




On 11/02/18 09:23, Paolo Monella wrote:

Dear all,

I'm tutoring a student for an MA thesis. She's creating TEI XML rhetorical
annotation on a Latin text, like this:

[in the teiHeader:]
                 <interp <a class="moz-txt-link-freetext" href="xml:id=">xml:id="anafora" sameAs=
"https://www.wikidata.
org/wiki/Q486050" <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.wikidata.org%2Fwiki%2FQ486050&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=k%2B4vRD%2FXWq9beqs4yqxdmJ1SoX1NjNZzRA%2BWG8QO9IA%3D&reserved=0> <https://www.wikidata.org/wiki/Q486050> <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.wikidata.org%2Fwiki%2FQ486050&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=k%2B4vRD%2FXWq9beqs4yqxdmJ1SoX1NjNZzRA%2BWG8QO9IA%3D&reserved=0>>Anafora</interp>


[in the body:]
        <lb n="123.17"/>'scripsit causas: quia, <seg ana="#anafora">cum
semel appellassem, cum <supplied>iterum appellassem,
        <lb n="123.18"/>cum</supplied> tertio appellassem</seg>, non
corrupi'. [etc.]

We're trying to also express these statemtents (passage xyx has rhetorical
figure: anaphora) in LOD. Do you have any suggestions?

I did some research, but I can't find hasRhetoricalFigure anywhere except
inhttp://www.repositori.com/sw/onto/RhetoricalDevices.owl <https://na01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.repositori.com%2Fsw%2Fonto%2FRhetoricalDevices.owl&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=OnrpazEDej4bqF8sGyF0u1kQxOJMpzeyFFyqdzBaykg%3D&reserved=0>

I'm new to LOD, and I'm afraid that the following code doesn't even make
sense:

<?xml version="1.0"?>
<rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" <https://na01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.w3.org%2F1999%2F02%2F22-rdf-syntax-ns%23&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=G7CdbsEK%2BgeLk3mZpAegalg5cYKEJJB0tWDmAYorz5E%3D&reserved=0><http://www.w3.org/1999/02/22-rdf-syntax-ns#> <https://na01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.w3.org%2F1999%2F02%2F22-rdf-syntax-ns%23&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=G7CdbsEK%2BgeLk3mZpAegalg5cYKEJJB0tWDmAYorz5E%3D&reserved=0>
  <rdf:Description rdf:about="annotated-document.xml#id-of-annotated-span">

    <rdf:instanceOf>
    https://www.wikidata.org/wiki/Q486050 <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.wikidata.org%2Fwiki%2FQ486050&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=k%2B4vRD%2FXWq9beqs4yqxdmJ1SoX1NjNZzRA%2BWG8QO9IA%3D&reserved=0>
    </rdf:instanceOf>
  </rdf:Description>
</rdf:RDF>

Thank you,
Paolo










--
Elisa Beshero-Bondar, PhD
Director, Center for the Digital Text | Associate Professor of English
University of Pittsburgh at Greensburg | Humanities Division
150 Finoli Drive
Greensburg, PA  15601  USA
E-mail:[hidden email]
Development site: http://newtfire.org

--
Richard Light
Reply | Threaded
Open this post in threaded view
|

Re: Rhetorical annotation and LOD

Elisa Beshero-Bondar
Hi Richard-- That all makes sense: participation in LOD may well strengthen its stability, and the use of the linkages should be enriching and meaningful and not just something we feel compelled to do. (Having a really well-defined and reasonably reliable reference point for anaphora should be helpful, and if multiple TEI projects connect with it, all benefit from shared association.)

As for the place in the TEI header, for classification systems, we've got <classDecl> which holds <taxonomy> : http://www.tei-c.org/release/doc/tei-p5-doc/en/html/ref-classDecl.html  , and we've got an open ticket to include a wider variety of examples: https://github.com/TEIC/TEI/issues/1640 
This is where I'd probably first look for making transformations to LOD, since often we're doing that point to external references.

Elisa



On Sun, Feb 11, 2018 at 12:27 PM, Richard Light <[hidden email]> wrote:

On 11/02/2018 16:13, Elisa Beshero-Bondar wrote:
Lou-- All in good "musing"--which is one thing we do well on this list! :-)  One thing I think our little debate is demonstrating is a larger tension between TEI and LOD. Each of these considers itself a system to encode information sustainably and interchangeably, and each is a little at odds with the other. TEI is not structured like RDF, but the relationships we encode can be transformed and expressed in those terms. Some have argued that TEI is "already" LOD, but practitioners of LOD complain that the values of pointers in our attributes need to be fully spelled out in each literal location for that to be achieved (on each and every seg element that has to be decoded to find its linked-open-data relationship). 

We can do that, but the LOD community isn't as stable as some of us would like it to be, and perhaps the expression and cataloging of differences is as lasting as we would hope of our linked networks of information. Some of us in the TEI world ind LOD a little uncertain and vexingly limited (as Paolo indicated--there's just one available vocabulary to work with, and maybe we're hoping it will be improved someday). We may want to participate in LOD following its vocabularies, but it might be that those will change, or more and better will become available. I don't like the sense that inventors of (not-so?) controlled vocabularies project a lasting stability to their structures, and so I guess I favor encodings that are a little more open to "mutability" in the LOD world. 

Anyway, for playing in a friendly way with LOD, and maintaining a sense of internal consistency within a project, I think the teiHeader is a good place to define relationships that might have to change later.
I would agree with that approach.  I have it in mind that a generic XSLT which publishes the TEI Header as a set of RDF assertions could be a generally useful addition to our armoury.  The development of such a transform would be greatly aided by community agreement on where and how LOD URLs might be included within the Header.

While it is valid to be concerned about the stability of LOD URLs, I wold argue that one way to help ensure that they do remain stable is to make it publicly known that they are in active use.  It's something of a chicken and egg situation.  I too have reservations about RDF, but I don't see any alternative to LOD URLs as a means of achieving semantic interoperability across the TEI world and beyond it.

Something else to bear in mind is that such URLs can be more than an additional recording burden and a source of worry: they can also provide additional information 'for free'.  In the case we are discussing:
https://www.wikidata.org/wiki/Q486050
you don't get a lot beyond multiple translations of 'anaphora' (although that might be useful).  Something I wrote to facilitate the use of LD resources in an XSLT processing environment is a little 'forwarder' program, which simply adds a suitable Accept header to an HTTP request, so that you can use the document() function on the URL and get RDF/XML back.

Best wishes,

Richard



Cheers,

Elisa


On Sun, Feb 11, 2018 at 11:10 AM, Lou Burnard <[hidden email]> wrote:

The argument that one might want to associate ANAPHORA with more than one definition is certainly more persuasive, though I don't know that such is the case here.

I wasn't making any criticism of this particular project, by the way; merely musing aloud about simpler ways of expressing a very simple idea.


On 11/02/18 16:03, Elisa Beshero-Bondar wrote:
And I am not "convinced" by your counterpoint, Lou. You seem to be
qualifying the project team's usage of "#anaphora" as somehow just as
meaningful as any random string of text. It isn't--its meaning, as Paolo
indicated, is defined in the teiHeader where such things are quite properly
defined.

That is, there isn't anything wrong with their current practice, and quite
a lot that's right about it. I'll even add something more: What if the
project developers wanted to add linkages to other rhetorical databanks to
their project, to encode multiple definitions of anaphora? Would you then
argue that each of those should literally appear in the @ana on each and
every seg?

Elisa

On Sun, Feb 11, 2018 at 10:57 AM, Lou Burnard [hidden email]
wrote:

Good point, but it doesn't convince me.

This project says "I want to call those things ANAPHORA (oh, and by
ANAPHORA I mean the same thing Wikidata calls Q486050)"

So if the meaning of "Wikidata Q486050" suddenly ceases to exist for some
reason, all I know is that this project considers these things to be the
same sort of thing (ANAPHORA) but I don't know what that means. In which
case, I might as well have called them "FRIEDTOMATOES" or even
"wikidata:W486050": I wouldn't be any better or worse off.


On 11/02/18 15:48, Elisa Beshero-Bondar wrote:

Hi Lou-- That raises the question of whether the project considers the
wikidata definitions to be permanently authoritative. If you've got those
pointers down in every seg, and the wiki goes away or changes its numbers,
well it's harder to rip them out later. If you've got the relationships
simply defined up the teiHeader, it's less code to have to change if the
wikidata goes away someday and the pointers need to resolve differently.

I'm not just raising that to be contrarian--That's a serious issue in the
LOD world! (Nought in this age is lasting...Nought endures but
mutability...) The TEI encoding of #anaphora on Paolo's project might be
best left on its own, and the rhetorical definitions left relatively easy
to redefine should the project developers decide they want to change the
links out or point to a better definitive resource.

Elisa



On Sun, Feb 11, 2018 at 10:20 AM, Lou Burnard [hidden email] [hidden email]
wrote:


I'm always happy to expose myself to ridicule on this list, so I would
like to know why the following wouldn't be easier and just as explicit:


   <lb n="123.17"/>'scripsit causas: quia, <seg ana="
https://www.wikidata.org/wiki/Q486050" <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.wikidata.org%2Fwiki%2FQ486050&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=k%2B4vRD%2FXWq9beqs4yqxdmJ1SoX1NjNZzRA%2BWG8QO9IA%3D&reserved=0>>cum semel appellassem, cum
<supplied>iterum appellassem,
        <lb n="123.18"/>cum</supplied> tertio appellassem</seg>, non
corrupi'. [etc.]

Or, of course, you could do a prefixdef for "https://wikidata.org/wiki" <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwikidata.org%2Fwiki&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=dRVPTdRfLncAJBXgSdjBR%2BpE0WgXVK0FBt2APHyhojE%3D&reserved=0><https://wikidata.org/wiki> <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwikidata.org%2Fwiki&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=dRVPTdRfLncAJBXgSdjBR%2BpE0WgXVK0FBt2APHyhojE%3D&reserved=0> and simply say

<seg ana="wikidata:Q486050">


L




On 11/02/18 09:23, Paolo Monella wrote:

Dear all,

I'm tutoring a student for an MA thesis. She's creating TEI XML rhetorical
annotation on a Latin text, like this:

[in the teiHeader:]
                 <interp xml:id="anafora" sameAs=
"https://www.wikidata.
org/wiki/Q486050" <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.wikidata.org%2Fwiki%2FQ486050&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=k%2B4vRD%2FXWq9beqs4yqxdmJ1SoX1NjNZzRA%2BWG8QO9IA%3D&reserved=0> <https://www.wikidata.org/wiki/Q486050> <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.wikidata.org%2Fwiki%2FQ486050&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=k%2B4vRD%2FXWq9beqs4yqxdmJ1SoX1NjNZzRA%2BWG8QO9IA%3D&reserved=0>>Anafora</interp>


[in the body:]
        <lb n="123.17"/>'scripsit causas: quia, <seg ana="#anafora">cum
semel appellassem, cum <supplied>iterum appellassem,
        <lb n="123.18"/>cum</supplied> tertio appellassem</seg>, non
corrupi'. [etc.]

We're trying to also express these statemtents (passage xyx has rhetorical
figure: anaphora) in LOD. Do you have any suggestions?

I did some research, but I can't find hasRhetoricalFigure anywhere except
inhttp://www.repositori.com/sw/onto/RhetoricalDevices.owl <https://na01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.repositori.com%2Fsw%2Fonto%2FRhetoricalDevices.owl&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=OnrpazEDej4bqF8sGyF0u1kQxOJMpzeyFFyqdzBaykg%3D&reserved=0>

I'm new to LOD, and I'm afraid that the following code doesn't even make
sense:

<?xml version="1.0"?>
<rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" <https://na01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.w3.org%2F1999%2F02%2F22-rdf-syntax-ns%23&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=G7CdbsEK%2BgeLk3mZpAegalg5cYKEJJB0tWDmAYorz5E%3D&reserved=0><http://www.w3.org/1999/02/22-rdf-syntax-ns#> <https://na01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.w3.org%2F1999%2F02%2F22-rdf-syntax-ns%23&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=G7CdbsEK%2BgeLk3mZpAegalg5cYKEJJB0tWDmAYorz5E%3D&reserved=0>
  <rdf:Description rdf:about="annotated-document.xml#id-of-annotated-span">

    <rdf:instanceOf>
    https://www.wikidata.org/wiki/Q486050 <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.wikidata.org%2Fwiki%2FQ486050&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=k%2B4vRD%2FXWq9beqs4yqxdmJ1SoX1NjNZzRA%2BWG8QO9IA%3D&reserved=0>
    </rdf:instanceOf>
  </rdf:Description>
</rdf:RDF>

Thank you,
Paolo










--
Elisa Beshero-Bondar, PhD
Director, Center for the Digital Text | Associate Professor of English
University of Pittsburgh at Greensburg | Humanities Division
150 Finoli Drive
Greensburg, PA  15601  USA
E-mail:[hidden email]
Development site: http://newtfire.org

--
Richard Light



--
Elisa Beshero-Bondar, PhD
Director, Center for the Digital Text | Associate Professor of English
University of Pittsburgh at Greensburg | Humanities Division
150 Finoli Drive
Greensburg, PA  15601  USA
E-mail:[hidden email]
Development site: http://newtfire.org
Reply | Threaded
Open this post in threaded view
|

Re: Rhetorical annotation and LOD

David Maus
In reply to this post by Paolo Monella
Hi Paolo,

On Sun, 11 Feb 2018 10:23:05 +0100,
Paolo Monella wrote:

>
> Dear all,
>
> I'm tutoring a student for an MA thesis. She's creating TEI XML
> rhetorical annotation on a Latin text, like this:
>
> [in the teiHeader:]
>                  <interp xml:id="anafora"
> sameAs="https://www.wikidata.org/wiki/Q486050">Anafora</interp>
>
> [in the body:]
>         <lb n="123.17"/>'scripsit causas: quia, <seg
> ana="#anafora">cum semel appellassem, cum <supplied>iterum
> appellassem,
>         <lb n="123.18"/>cum</supplied> tertio appellassem</seg>, non
> corrupi'. [etc.]
>
> We're trying to also express these statemtents (passage xyx has
> rhetorical figure: anaphora) in LOD. Do you have any suggestions?
>
> I did some research, but I can't find hasRhetoricalFigure anywhere except in
> http://www.repositori.com/sw/onto/RhetoricalDevices.owl

What works best for me in such cases is a two step process: First I
define my own ontology with my own classes and properties. This step
is important because it helps me to focus on modelling the problem
domain. In a second step I link my ontology to one or more target
ontologies. Or I replace all of my classes/properties with those of
one or more target ontologies.

>
> I'm new to LOD, and I'm afraid that the following code doesn't even
> make sense:
>
> <?xml version="1.0"?>
> <rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
>   <rdf:Description rdf:about="annotated-document.xml#id-of-annotated-span">
>     <rdf:instanceOf>
>     https://www.wikidata.org/wiki/Q486050
>     </rdf:instanceOf>
>   </rdf:Description>
> </rdf:RDF>
>

This should be written as:

<rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
  <rdf:Description rdf:about="annotated-document.xml#id-of-annotated-span">
    <rdf:type rdf:resource="https://www.wikidata.org/wiki/Q486050"/>
  </rdf:Description>
</rdf:RDF>

Regarding the construction of the URIs and LOD principles I found the
following books very helpful:

Tom Heath, und Christian Bizer. 2011. Linked Data. Evolving the Web
into a Global Data Space. 1. Aufl. Bd. 1. Synthesis Lectures on the
Semantic Web: Theory and Technology. Morgan & Claypool.

http://linkeddatabook.com/editions/1.0/

Powers, Shelley. 2003. Practical RDF: Solving Problems with the
Resource Description Framework. Bejing/Cambridge/Farnham et al.:
O’Reilly Media.

Hope that helps,
  -- David

> Thank you,
> Paolo

--
David Maus, Bibliothekarische IT / Digital Humanities
Herzog August Bibliothek, D-38299 Wolfenbüttel, Phone +49 5331 808-317
PGP Key 0x292EB8E568C7E0C9
http://dmaus.name ~ http://github.com/dmj
Reply | Threaded
Open this post in threaded view
|

Re: Rhetorical annotation and LOD

Conal Tuohy-3
One other little detail is that you should not use the URI <https://www.wikidata.org/wiki/Q486050> to denote the rhetorical figure called "anaphora". That URI actually denotes a web page on the topic of anaphora; it does NOT itself denote the concept of anaphora. The correct URI to use to denote the concept is actually <http://www.wikidata.org/entity/Q486050>. You can find this URI hidden rather unhelpfully in the left margin of the wikidata page, labelled "concept URI".

So I believe a correct RDF encoding would be:

<rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
  <rdf:Description rdf:about="annotated-document.xml#id-of-annotated-span">
    <rdf:type rdf:resource="http://www.wikidata.org/entity/Q486050"/>
  </rdf:Description>
</rdf:RDF>

The reason for the dual URIs (which is common practice in LOD) is that it's necessary to be able to make statements in RDF about things which are information resources on the web, as well as things which are NOT information resources on the web. In the case that a URI denotes a thing which is NOT an information resource, normally the web server responsible for that URI will return a redirect to a different URI (in effect saying "anaphora is not an information resource, but please see this other page which contains information about it". You will note that if you access <http://www.wikidata.org/entity/Q486050> in a browser, you will be redirected automatically to <https://www.wikidata.org/wiki/Q486050>

Conal

On 12 February 2018 at 17:56, David Maus <[hidden email]> wrote:
Hi Paolo,

On Sun, 11 Feb 2018 10:23:05 +0100,
Paolo Monella wrote:
>
> Dear all,
>
> I'm tutoring a student for an MA thesis. She's creating TEI XML
> rhetorical annotation on a Latin text, like this:
>
> [in the teiHeader:]
>                  <interp xml:id="anafora"
> sameAs="https://www.wikidata.org/wiki/Q486050">Anafora</interp>
>
> [in the body:]
>         <lb n="123.17"/>'scripsit causas: quia, <seg
> ana="#anafora">cum semel appellassem, cum <supplied>iterum
> appellassem,
>         <lb n="123.18"/>cum</supplied> tertio appellassem</seg>, non
> corrupi'. [etc.]
>
> We're trying to also express these statemtents (passage xyx has
> rhetorical figure: anaphora) in LOD. Do you have any suggestions?
>
> I did some research, but I can't find hasRhetoricalFigure anywhere except in
> http://www.repositori.com/sw/onto/RhetoricalDevices.owl

What works best for me in such cases is a two step process: First I
define my own ontology with my own classes and properties. This step
is important because it helps me to focus on modelling the problem
domain. In a second step I link my ontology to one or more target
ontologies. Or I replace all of my classes/properties with those of
one or more target ontologies.

>
> I'm new to LOD, and I'm afraid that the following code doesn't even
> make sense:
>
> <?xml version="1.0"?>
> <rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
>   <rdf:Description rdf:about="annotated-document.xml#id-of-annotated-span">
>     <rdf:instanceOf>
>     https://www.wikidata.org/wiki/Q486050
>     </rdf:instanceOf>
>   </rdf:Description>
> </rdf:RDF>
>

This should be written as:

<rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
  <rdf:Description rdf:about="annotated-document.xml#id-of-annotated-span">
    <rdf:type rdf:resource="https://www.wikidata.org/wiki/Q486050"/>
  </rdf:Description>
</rdf:RDF>

Regarding the construction of the URIs and LOD principles I found the
following books very helpful:

Tom Heath, und Christian Bizer. 2011. Linked Data. Evolving the Web
into a Global Data Space. 1. Aufl. Bd. 1. Synthesis Lectures on the
Semantic Web: Theory and Technology. Morgan & Claypool.

http://linkeddatabook.com/editions/1.0/

Powers, Shelley. 2003. Practical RDF: Solving Problems with the
Resource Description Framework. Bejing/Cambridge/Farnham et al.:
O’Reilly Media.

Hope that helps,
  -- David

> Thank you,
> Paolo

--
David Maus, Bibliothekarische IT / Digital Humanities
Herzog August Bibliothek, D-38299 Wolfenbüttel, Phone <a href="tel:%2B49%205331%20808-317" value="+495331808317">+49 5331 808-317
PGP Key 0x292EB8E568C7E0C9
http://dmaus.name ~ http://github.com/dmj



--
@conal_tuohy
+61-466-324297
Reply | Threaded
Open this post in threaded view
|

Re: Rhetorical annotation and LOD

Paolo Monella
Thank you David and Conal: I'll use your RDF encoding. I also appreciate
David's bibliographical references and Conan's explanation of the
resource/entity difference.
Best,
Paolo

Il 12/02/2018 09:55, Conal Tuohy ha scritto:

> One other little detail is that you should not use the URI
> <https://www.wikidata.org/wiki/Q486050 
> <https://www.wikidata.org/wiki/Q486050>> to denote the rhetorical figure
> called "anaphora". That URI actually denotes a web page on the topic of
> anaphora; it does NOT itself denote the concept of anaphora. The correct
> URI to use to denote the concept is actually
> <http://www.wikidata.org/entity/Q486050>. You can find this URI hidden
> rather unhelpfully in the left margin of the wikidata page, labelled
> "concept URI".
>
> So I believe a correct RDF encoding would be:
>
> <rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns# 
> <http://www.w3.org/1999/02/22-rdf-syntax-ns#>"
>    <rdf:Description rdf:about="annotated-document.xml#id-of-annotated-span">
>      <rdf:type rdf:resource="http://www.wikidata.org/entity/Q486050"/>
>    </rdf:Description>
> </rdf:RDF>
>
> The reason for the dual URIs (which is common practice in LOD) is that
> it's necessary to be able to make statements in RDF about things which
> are information resources on the web, as well as things which are NOT
> information resources on the web. In the case that a URI denotes a thing
> which is NOT an information resource, normally the web server
> responsible for that URI will return a redirect to a different URI (in
> effect saying "anaphora is not an information resource, but please see
> this other page which contains information about it". You will note that
> if you access <http://www.wikidata.org/entity/Q486050> in a browser, you
> will be redirected automatically to
> <https://www.wikidata.org/wiki/Q486050 
> <https://www.wikidata.org/wiki/Q486050>>
>
> Conal
>
> On 12 February 2018 at 17:56, David Maus <[hidden email]
> <mailto:[hidden email]>> wrote:
>
>     Hi Paolo,
>
>     On Sun, 11 Feb 2018 10:23:05 +0100,
>     Paolo Monella wrote:
>     >
>     > Dear all,
>     >
>     > I'm tutoring a student for an MA thesis. She's creating TEI XML
>     > rhetorical annotation on a Latin text, like this:
>     >
>     > [in the teiHeader:]
>     >                  <interp xml:id="anafora"
>     > sameAs="https://www.wikidata.org/wiki/Q486050
>     <https://www.wikidata.org/wiki/Q486050>">Anafora</interp>
>     >
>     > [in the body:]
>     >         <lb n="123.17"/>'scripsit causas: quia, <seg
>     > ana="#anafora">cum semel appellassem, cum <supplied>iterum
>     > appellassem,
>     >         <lb n="123.18"/>cum</supplied> tertio appellassem</seg>, non
>     > corrupi'. [etc.]
>     >
>     > We're trying to also express these statemtents (passage xyx has
>     > rhetorical figure: anaphora) in LOD. Do you have any suggestions?
>     >
>     > I did some research, but I can't find hasRhetoricalFigure anywhere except in
>     > http://www.repositori.com/sw/onto/RhetoricalDevices.owl
>     <http://www.repositori.com/sw/onto/RhetoricalDevices.owl>
>
>     What works best for me in such cases is a two step process: First I
>     define my own ontology with my own classes and properties. This step
>     is important because it helps me to focus on modelling the problem
>     domain. In a second step I link my ontology to one or more target
>     ontologies. Or I replace all of my classes/properties with those of
>     one or more target ontologies.
>
>     >
>     > I'm new to LOD, and I'm afraid that the following code doesn't even
>     > make sense:
>     >
>     > <?xml version="1.0"?>
>     > <rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#
>     <http://www.w3.org/1999/02/22-rdf-syntax-ns#>"
>     >   <rdf:Description rdf:about="annotated-document.xml#id-of-annotated-span">
>     >     <rdf:instanceOf>
>     >     https://www.wikidata.org/wiki/Q486050
>     <https://www.wikidata.org/wiki/Q486050>
>     >     </rdf:instanceOf>
>     >   </rdf:Description>
>     > </rdf:RDF>
>     >
>
>     This should be written as:
>
>     <rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#
>     <http://www.w3.org/1999/02/22-rdf-syntax-ns#>"
>        <rdf:Description
>     rdf:about="annotated-document.xml#id-of-annotated-span">
>          <rdf:type rdf:resource="https://www.wikidata.org/wiki/Q486050
>     <https://www.wikidata.org/wiki/Q486050>"/>
>        </rdf:Description>
>     </rdf:RDF>
>
>     Regarding the construction of the URIs and LOD principles I found the
>     following books very helpful:
>
>     Tom Heath, und Christian Bizer. 2011. Linked Data. Evolving the Web
>     into a Global Data Space. 1. Aufl. Bd. 1. Synthesis Lectures on the
>     Semantic Web: Theory and Technology. Morgan & Claypool.
>
>     http://linkeddatabook.com/editions/1.0/
>     <http://linkeddatabook.com/editions/1.0/>
>
>     Powers, Shelley. 2003. Practical RDF: Solving Problems with the
>     Resource Description Framework. Bejing/Cambridge/Farnham et al.:
>     O’Reilly Media.
>
>     Hope that helps,
>        -- David
>
>      > Thank you,
>      > Paolo
>
>     --
>     David Maus, Bibliothekarische IT / Digital Humanities
>     Herzog August Bibliothek, D-38299 Wolfenbüttel, Phone +49 5331
>     808-317 <tel:%2B49%205331%20808-317>
>     PGP Key 0x292EB8E568C7E0C9
>     http://dmaus.name ~ http://github.com/dmj
>
>
>
>
> --
> Conal Tuohy
> http://conaltuohy.com/
> @conal_tuohy
> +61-466-324297
Reply | Threaded
Open this post in threaded view
|

Re: Rhetorical annotation and LOD

Paolo Monella
In reply to this post by Elisa Beshero-Bondar
This is one of the reasons why I like the TEI-L: you get both useful
practical advice and interesting general discussion.

I would use Lou's advice (with prefixDef):
1. Define a prefixDef for "https://wikidata.org/entity"
2. say <seg ana="wikidata:Q486050">

but I agree with Elisa that keeping my extra step in the teiHeader
(where I explicitly provide at least a name for each rhetorical figure)
is safer for durability, should the WikiData project not endure time:

<interp xml:id="anafora"
sameAs="https://www.wikidata.org/entity/Q486050">Anafora</interp>
<interp xml:id="entimema"
sameAs="https://www.wikidata.org/entity/Q422913">Entimema</interp>
...

It's only a few extra code lines, easily readable and mantainable by
other projects, so I think that they are a fair price to pay for durability.

In our small MA thesis, I don't think we'll need <classDecl>/<taxonomy>:
all we have is a list of <interp>s with rhetorical figure names. But
I'll keep it in mind for future projects, as well as David's two-step
process (modelling, then linking).

Thank you all!
Best,
Paolo


Il 11/02/2018 18:59, Elisa Beshero-Bondar ha scritto:

> Hi Richard-- That all makes sense: participation in LOD may well
> strengthen its stability, and the use of the linkages should be
> enriching and meaningful and not just something we feel compelled to do.
> (Having a really well-defined and reasonably reliable reference point
> for anaphora should be helpful, and if multiple TEI projects connect
> with it, all benefit from shared association.)
>
> As for the place in the TEI header, for classification systems, we've
> got <classDecl> which holds <taxonomy> :
> http://www.tei-c.org/release/doc/tei-p5-doc/en/html/ref-classDecl.html 
> , and we've got an open ticket to include a wider variety of examples:
> https://github.com/TEIC/TEI/issues/1640
> This is where I'd probably first look for making transformations to LOD,
> since often we're doing that point to external references.
>
> Elisa
>
>
>
> On Sun, Feb 11, 2018 at 12:27 PM, Richard Light
> <[hidden email] <mailto:[hidden email]>> wrote:
>
>
>     On 11/02/2018 16:13, Elisa Beshero-Bondar wrote:
>>     Lou-- All in good "musing"--which is one thing we do well on this
>>     list! :-)  One thing I think our little debate is demonstrating is
>>     a larger tension between TEI and LOD. Each of these considers
>>     itself a system to encode information sustainably and
>>     interchangeably, and each is a little at odds with the other. TEI
>>     is not structured like RDF, but the relationships we encode can be
>>     transformed and expressed in those terms. Some have argued that
>>     TEI is "already" LOD, but practitioners of LOD complain that the
>>     values of pointers in our attributes need to be fully spelled out
>>     in each literal location for that to be achieved (on each and
>>     every seg element that has to be decoded to find its
>>     linked-open-data relationship).
>>
>>     We can do that, but the LOD community isn't as stable as some of
>>     us would like it to be, and perhaps the expression and cataloging
>>     of differences is as lasting as we would hope of our linked
>>     networks of information. Some of us in the TEI world ind LOD a
>>     little uncertain and vexingly limited (as Paolo indicated--there's
>>     just one available vocabulary to work with, and maybe we're hoping
>>     it will be improved someday). We may want to participate in LOD
>>     following its vocabularies, but it might be that those will
>>     change, or more and better will become available. I don't like the
>>     sense that inventors of (not-so?) controlled vocabularies project
>>     a lasting stability to their structures, and so I guess I favor
>>     encodings that are a little more open to "mutability" in the LOD
>>     world.
>>
>>     Anyway, for playing in a friendly way with LOD, and maintaining a
>>     sense of internal consistency within a project, I think the
>>     teiHeader is a good place to define relationships that might have
>>     to change later.
>     I would agree with that approach.  I have it in mind that a generic
>     XSLT which publishes the TEI Header as a set of RDF assertions could
>     be a generally useful addition to our armoury.  The development of
>     such a transform would be greatly aided by community agreement on
>     where and how LOD URLs might be included within the Header.
>
>     While it is valid to be concerned about the stability of LOD URLs, I
>     wold argue that one way to help ensure that they do remain stable is
>     to make it publicly known that they are in active use.  It's
>     something of a chicken and egg situation.  I too have reservations
>     about RDF, but I don't see any alternative to LOD URLs as a means of
>     achieving semantic interoperability across the TEI world and beyond it.
>
>     Something else to bear in mind is that such URLs can be more than an
>     additional recording burden and a source of worry: they can also
>     provide additional information 'for free'.  In the case we are
>     discussing:
>
>     https://www.wikidata.org/wiki/Q486050
>     <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.wikidata.org%2Fwiki%2FQ486050&data=01%7C01%7Cebb8%40pitt.edu%7Cb68450f3699d42e9f10e08d57169f866%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=frWh6hnnaj4xOWrSUgVHtSmENj0w6c4reJKFr%2B30qHU%3D&reserved=0>
>
>     you don't get a lot beyond multiple translations of 'anaphora'
>     (although that might be useful).  Something I wrote to facilitate
>     the use of LD resources in an XSLT processing environment is a
>     little 'forwarder' program, which simply adds a suitable Accept
>     header to an HTTP request, so that you can use the document()
>     function on the URL and get RDF/XML back.
>
>     Best wishes,
>
>     Richard
>
>
>>
>>     Cheers,
>>
>>     Elisa
>>
>>
>>     On Sun, Feb 11, 2018 at 11:10 AM, Lou Burnard
>>     <[hidden email]
>>     <mailto:[hidden email]>> wrote:
>>
>>         The argument that one might want to associate ANAPHORA with
>>         more than one definition is certainly more persuasive, though
>>         I don't know that such is the case here.
>>
>>         I wasn't making any criticism of this particular project, by
>>         the way; merely musing aloud about simpler ways of expressing
>>         a very simple idea.
>>
>>
>>         On 11/02/18 16:03, Elisa Beshero-Bondar wrote:
>>>         And I am not "convinced" by your counterpoint, Lou. You seem to be
>>>         qualifying the project team's usage of "#anaphora" as somehow just as
>>>         meaningful as any random string of text. It isn't--its meaning, as Paolo
>>>         indicated, is defined in the teiHeader where such things are quite properly
>>>         defined.
>>>
>>>         That is, there isn't anything wrong with their current practice, and quite
>>>         a lot that's right about it. I'll even add something more: What if the
>>>         project developers wanted to add linkages to other rhetorical databanks to
>>>         their project, to encode multiple definitions of anaphora? Would you then
>>>         argue that each of those should literally appear in the @ana on each and
>>>         every seg?
>>>
>>>         Elisa
>>>
>>>         On Sun, Feb 11, 2018 at 10:57 AM, Lou Burnard<[hidden email]>
>>>         <mailto:[hidden email]>
>>>         wrote:
>>>
>>>>         Good point, but it doesn't convince me.
>>>>
>>>>         This project says "I want to call those things ANAPHORA (oh, and by
>>>>         ANAPHORA I mean the same thing Wikidata calls Q486050)"
>>>>
>>>>         So if the meaning of "Wikidata Q486050" suddenly ceases to exist for some
>>>>         reason, all I know is that this project considers these things to be the
>>>>         same sort of thing (ANAPHORA) but I don't know what that means. In which
>>>>         case, I might as well have called them "FRIEDTOMATOES" or even
>>>>         "wikidata:W486050": I wouldn't be any better or worse off.
>>>>
>>>>
>>>>         On 11/02/18 15:48, Elisa Beshero-Bondar wrote:
>>>>
>>>>         Hi Lou-- That raises the question of whether the project considers the
>>>>         wikidata definitions to be permanently authoritative. If you've got those
>>>>         pointers down in every seg, and the wiki goes away or changes its numbers,
>>>>         well it's harder to rip them out later. If you've got the relationships
>>>>         simply defined up the teiHeader, it's less code to have to change if the
>>>>         wikidata goes away someday and the pointers need to resolve differently.
>>>>
>>>>         I'm not just raising that to be contrarian--That's a serious issue in the
>>>>         LOD world! (Nought in this age is lasting...Nought endures but
>>>>         mutability...) The TEI encoding of #anaphora on Paolo's project might be
>>>>         best left on its own, and the rhetorical definitions left relatively easy
>>>>         to redefine should the project developers decide they want to change the
>>>>         links out or point to a better definitive resource.
>>>>
>>>>         Elisa
>>>>
>>>>
>>>>
>>>>         On Sun, Feb 11, 2018 at 10:20 AM, Lou Burnard<[hidden email]>
>>>>         <mailto:[hidden email]>  <[hidden email]>
>>>>         <mailto:[hidden email]>wrote: I'm always happy
>>>>         to expose myself to ridicule on this list, so I would like
>>>>         to know why the following wouldn't be easier and just as
>>>>         explicit: <lb n="123.17"/>'scripsit causas: quia, <seg ana="
>>>>         https://www.wikidata.org/wiki/Q486050"
>>>>         <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.wikidata.org%2Fwiki%2FQ486050&data=01%7C01%7Cebb8%40pitt.edu%7Cb68450f3699d42e9f10e08d57169f866%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=frWh6hnnaj4xOWrSUgVHtSmENj0w6c4reJKFr%2B30qHU%3D&reserved=0>  <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.wikidata.org%2Fwiki%2FQ486050&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=k%2B4vRD%2FXWq9beqs4yqxdmJ1SoX1NjNZzRA%2BWG8QO9IA%3D&reserved=0>
>>>>         <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.wikidata.org%2Fwiki%2FQ486050&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=k%2B4vRD%2FXWq9beqs4yqxdmJ1SoX1NjNZzRA%2BWG8QO9IA%3D&reserved=0>>cum semel appellassem, cum
>>>>         <supplied>iterum appellassem,
>>>>                  <lb n="123.18"/>cum</supplied> tertio appellassem</seg>, non
>>>>         corrupi'. [etc.]
>>>>
>>>>         Or, of course, you could do a prefixdef for"https://wikidata.org/wiki"
>>>>         <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwikidata.org%2Fwiki&data=01%7C01%7Cebb8%40pitt.edu%7Cb68450f3699d42e9f10e08d57169f866%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=%2BB6hsn1FMdHx1RqR27CTaGN%2BrWsYlpwsTX8RtZre3Jg%3D&reserved=0>  <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwikidata.org%2Fwiki&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=dRVPTdRfLncAJBXgSdjBR%2BpE0WgXVK0FBt2APHyhojE%3D&reserved=0>
>>>>         <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwikidata.org%2Fwiki&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=dRVPTdRfLncAJBXgSdjBR%2BpE0WgXVK0FBt2APHyhojE%3D&reserved=0><https://wikidata.org/wiki>
>>>>         <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwikidata.org%2Fwiki&data=01%7C01%7Cebb8%40pitt.edu%7Cb68450f3699d42e9f10e08d57169f866%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=%2BB6hsn1FMdHx1RqR27CTaGN%2BrWsYlpwsTX8RtZre3Jg%3D&reserved=0>  <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwikidata.org%2Fwiki&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=dRVPTdRfLncAJBXgSdjBR%2BpE0WgXVK0FBt2APHyhojE%3D&reserved=0>
>>>>         <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwikidata.org%2Fwiki&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=dRVPTdRfLncAJBXgSdjBR%2BpE0WgXVK0FBt2APHyhojE%3D&reserved=0>  and simply say
>>>>
>>>>         <seg ana="wikidata:Q486050">
>>>>
>>>>
>>>>         L
>>>>
>>>>
>>>>
>>>>
>>>>         On 11/02/18 09:23, Paolo Monella wrote:
>>>>
>>>>         Dear all,
>>>>
>>>>         I'm tutoring a student for an MA thesis. She's creating TEI XML rhetorical
>>>>         annotation on a Latin text, like this:
>>>>
>>>>         [in the teiHeader:]
>>>>                           <interpxml:id="anafora" sameAs=
>>>>         "https://www.wikidata. org/wiki/Q486050"
>>>>         <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.wikidata.org%2Fwiki%2FQ486050&data=01%7C01%7Cebb8%40pitt.edu%7Cb68450f3699d42e9f10e08d57169f866%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=frWh6hnnaj4xOWrSUgVHtSmENj0w6c4reJKFr%2B30qHU%3D&reserved=0>  <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.wikidata.org%2Fwiki%2FQ486050&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=k%2B4vRD%2FXWq9beqs4yqxdmJ1SoX1NjNZzRA%2BWG8QO9IA%3D&reserved=0>
>>>>         <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.wikidata.org%2Fwiki%2FQ486050&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=k%2B4vRD%2FXWq9beqs4yqxdmJ1SoX1NjNZzRA%2BWG8QO9IA%3D&reserved=0>  <https://www.wikidata.org/wiki/Q486050>
>>>>         <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.wikidata.org%2Fwiki%2FQ486050&data=01%7C01%7Cebb8%40pitt.edu%7Cb68450f3699d42e9f10e08d57169f866%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=frWh6hnnaj4xOWrSUgVHtSmENj0w6c4reJKFr%2B30qHU%3D&reserved=0>  <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.wikidata.org%2Fwiki%2FQ486050&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=k%2B4vRD%2FXWq9beqs4yqxdmJ1SoX1NjNZzRA%2BWG8QO9IA%3D&reserved=0>
>>>>         <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.wikidata.org%2Fwiki%2FQ486050&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=k%2B4vRD%2FXWq9beqs4yqxdmJ1SoX1NjNZzRA%2BWG8QO9IA%3D&reserved=0>>Anafora</interp>
>>>>
>>>>
>>>>         [in the body:]
>>>>                  <lb n="123.17"/>'scripsit causas: quia, <seg ana="#anafora">cum
>>>>         semel appellassem, cum <supplied>iterum appellassem,
>>>>                  <lb n="123.18"/>cum</supplied> tertio appellassem</seg>, non
>>>>         corrupi'. [etc.]
>>>>
>>>>         We're trying to also express these statemtents (passage xyx has rhetorical
>>>>         figure: anaphora) in LOD. Do you have any suggestions?
>>>>
>>>>         I did some research, but I can't find hasRhetoricalFigure anywhere except
>>>>         inhttp://www.repositori.com/sw/onto/RhetoricalDevices.owl
>>>>         <http://www.repositori.com/sw/onto/RhetoricalDevices.owl>  <https://na01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.repositori.com%2Fsw%2Fonto%2FRhetoricalDevices.owl&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=OnrpazEDej4bqF8sGyF0u1kQxOJMpzeyFFyqdzBaykg%3D&reserved=0>
>>>>         <https://na01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.repositori.com%2Fsw%2Fonto%2FRhetoricalDevices.owl&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=OnrpazEDej4bqF8sGyF0u1kQxOJMpzeyFFyqdzBaykg%3D&reserved=0>
>>>>
>>>>         I'm new to LOD, and I'm afraid that the following code doesn't even make
>>>>         sense:
>>>>
>>>>         <?xml version="1.0"?>
>>>>         <rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
>>>>         <https://na01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.w3.org%2F1999%2F02%2F22-rdf-syntax-ns%23&data=01%7C01%7Cebb8%40pitt.edu%7Cb68450f3699d42e9f10e08d57169f866%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=wHqWhY2dfkwufphRXbPD7yGZvXHWq28cMHtW9ak%2FSKk%3D&reserved=0>  <https://na01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.w3.org%2F1999%2F02%2F22-rdf-syntax-ns%23&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=G7CdbsEK%2BgeLk3mZpAegalg5cYKEJJB0tWDmAYorz5E%3D&reserved=0>
>>>>         <https://na01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.w3.org%2F1999%2F02%2F22-rdf-syntax-ns%23&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=G7CdbsEK%2BgeLk3mZpAegalg5cYKEJJB0tWDmAYorz5E%3D&reserved=0><http://www.w3.org/1999/02/22-rdf-syntax-ns#>
>>>>         <https://na01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.w3.org%2F1999%2F02%2F22-rdf-syntax-ns%23&data=01%7C01%7Cebb8%40pitt.edu%7Cb68450f3699d42e9f10e08d57169f866%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=wHqWhY2dfkwufphRXbPD7yGZvXHWq28cMHtW9ak%2FSKk%3D&reserved=0>  <https://na01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.w3.org%2F1999%2F02%2F22-rdf-syntax-ns%23&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=G7CdbsEK%2BgeLk3mZpAegalg5cYKEJJB0tWDmAYorz5E%3D&reserved=0>
>>>>         <https://na01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.w3.org%2F1999%2F02%2F22-rdf-syntax-ns%23&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=G7CdbsEK%2BgeLk3mZpAegalg5cYKEJJB0tWDmAYorz5E%3D&reserved=0><rdf:Description
>>>>         rdf:about="annotated-document.xml#id-of-annotated-span">
>>>>
>>>>              <rdf:instanceOf>
>>>>              https://www.wikidata.org/wiki/Q486050
>>>>         <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.wikidata.org%2Fwiki%2FQ486050&data=01%7C01%7Cebb8%40pitt.edu%7Cb68450f3699d42e9f10e08d57169f866%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=frWh6hnnaj4xOWrSUgVHtSmENj0w6c4reJKFr%2B30qHU%3D&reserved=0>  <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.wikidata.org%2Fwiki%2FQ486050&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=k%2B4vRD%2FXWq9beqs4yqxdmJ1SoX1NjNZzRA%2BWG8QO9IA%3D&reserved=0>
>>>>         <https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.wikidata.org%2Fwiki%2FQ486050&data=01%7C01%7Cebb8%40pitt.edu%7C4f6425d2e3d640c3cb8708d571682e73%7C9ef9f489e0a04eeb87cc3a526112fd0d%7C1&sdata=k%2B4vRD%2FXWq9beqs4yqxdmJ1SoX1NjNZzRA%2BWG8QO9IA%3D&reserved=0></rdf:instanceOf>
>>>>         </rdf:Description> </rdf:RDF> Thank you, Paolo
>>
>>
>>
>>
>>     --
>>     Elisa Beshero-Bondar, PhD
>>     Director, Center for the Digital Text | Associate Professor of English
>>     University of Pittsburgh at Greensburg | Humanities Division
>>     150 Finoli Drive
>>     <https://maps.google.com/?q=150+Finoli+Drive+%0D+Greensburg,+PA+%C2%A015601+%C2%A0USA&entry=gmail&source=g>
>>     Greensburg, PA  15601  USA
>>     E-mail: [hidden email] <mailto:[hidden email]>
>>     Development site: http://newtfire.org <http://newtfire.org/>
>
>     --
>     *Richard Light*
>
>
>
>
> --
> Elisa Beshero-Bondar, PhD
> Director, Center for the Digital Text | Associate Professor of English
> University of Pittsburgh at Greensburg | Humanities Division
> 150 Finoli Drive
> Greensburg, PA  15601  USA
> E-mail: [hidden email] <mailto:[hidden email]>
> Development site: http://newtfire.org <http://newtfire.org/>
Reply | Threaded
Open this post in threaded view
|

Re: Rhetorical annotation and LOD

Paolo Monella
In reply to this post by Conal Tuohy-3
Il 12/02/2018 09:55, Conal Tuohy ha scritto:
> <rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns# 
> <http://www.w3.org/1999/02/22-rdf-syntax-ns#>"
>    <rdf:Description rdf:about="annotated-document.xml#id-of-annotated-span">
>      <rdf:type rdf:resource="http://www.wikidata.org/entity/Q486050"/>
>    </rdf:Description>
> </rdf:RDF>

A final question: is there a more specific way to say (in RDF) that
http://www.wikidata.org/entity/Q486050 is not only the "type" of the
annotated span, but specifically its "rhetorical figure"? I'd prefer not
to define my own ontology. May I just refer to the "hasRhetoricalFigure"
property in http://www.repositori.com/sw/onto/RhetoricalDevices.owl ?
Thank you,
Paolo
Reply | Threaded
Open this post in threaded view
|

Re: Rhetorical annotation and LOD

David Maus
On Tue, 13 Feb 2018 12:46:46 +0100,
Paolo Monella wrote:

>
> Il 12/02/2018 09:55, Conal Tuohy ha scritto:
> > <rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#
> > <http://www.w3.org/1999/02/22-rdf-syntax-ns#>"
> >    <rdf:Description rdf:about="annotated-document.xml#id-of-annotated-span">
> >      <rdf:type rdf:resource="http://www.wikidata.org/entity/Q486050"/>
> >    </rdf:Description>
> > </rdf:RDF>
>
> A final question: is there a more specific way to say (in RDF) that
> http://www.wikidata.org/entity/Q486050 is not only the "type" of the
> annotated span, but specifically its "rhetorical figure"? I'd prefer
> not to define my own ontology. May I just refer to the
> "hasRhetoricalFigure" property in
> http://www.repositori.com/sw/onto/RhetoricalDevices.owl ?

I think it depends on the statement you want to make:

a) annotated-document.xml#id-of-annotated-span *is a* anaphora

use rdf:type

b) annotated-document.xml#id-of-annotated-span *has a* anaphora

use hasRhetoricalFigure and point to something with a rdf:type
http://www.wikidata.org/entity/Q486050

Rule of thumb: Use option b) if you need to make a distinction between
the annotated segment of text and its rhetorical figure, i.e. if you
view them as distinct entities.

HTH,
  -- David

--
David Maus, Bibliothekarische IT / Digital Humanities
Herzog August Bibliothek, D-38299 Wolfenbüttel, Phone +49 5331 808-317
PGP Key 0x292EB8E568C7E0C9
http://dmaus.name ~ http://github.com/dmj