?? how to model collection of works

[sdm10012]

using the Tolkein trilogy instance .... what's the convention for
modeling the parent / child subtype relationships of

Lord of the Rings

The Fellowship of the Ring
The Two Towers
The Return of the King

I'm assuming LOR would be parent "Context" and the book instances
would also be contexts ...

But should they all be "wrapped" in a collection ???

Enough of my speculation .... there must be a standard model for this
type of thing .... but I've exhausted my resourcefulness looking for a
contextus example.

thx,

Steve

[Faith]

On Mar 10, 7:15 pm, sdm10012 <steve.mi...@gmail.com> wrote:
> using the Tolkein trilogy instance .... what's the convention for
> modeling the parent / child subtype relationships of
>
> Lord of the Rings
>
>    The Fellowship of the Ring
>    The Two Towers
>    The Return of the King
>
> I'm assuming LOR would be parent "Context" and the book instances
> would also be contexts ...

Yep. And the 'book' contexts would exist within the parent LotR
context.

> But should they all be "wrapped" in a collection ???

I don't think you need to wrap it in a collection - the basic parent-
child relationship would be

TheFellowshipOfTheRing ome:exists-in TheLordOfTheRings

TheLordOfTheRings ome:allows-existence-of TheFellowshipOfTheRing etc.

(where TheFellowshipOfTheRing and TheLordOfTheRings are ome:Contexts)

The fun comes if you want to relate the sibling contexts to each other
- at that point you are basically playing around with timelines. I
don't think (and one of the others might correct me here) that there
is a way to directly relate the contexts to each other temporally
(there are obviously shadow-of, spin-off-of, etc relationships) - it's
all about where the events existing in each context take place on the
shared timeline. You could - if you really wanted - create a timeline
specifically for linking the contexts together with events that relate
to the start and end of each contextual time period but in that case
you better hope that there is no overlap between the time periods
covered by the contexts or you are likely to start contradicting
yourself when you do put the occurrences of events on the timeline (or
timelines). I believe the decision was that it was better to infer
relative position of the contexts whether, chronologically or
temporally or in what ever other way, based on how the events within
those contexts stacked up.

I hope that makes sense.

And if your next question is what is the property that links a
timeline to a Context (or any other expression) - I think we just used
the 'is' property (very philosophical: a thing is the same as the sum
of it's existence).

> Enough of my speculation .... there must be a standard model for this
> type of thing .... but I've exhausted my resourcefulness looking for a
> contextus example.

I do keep meaning to do a page of examples of contexts and
intertextuality and just haven't found the time sadly. I will get
around to it soon - honest. In case it is helpful have a look at the
examples at http://interaction.ecs.soton.ac.uk/ir/projects/ontomedia/examples/
These were *very* early examples that we worked on (so early I'm a
little surprised that they are still around online) so you will
notice, for example, that all of the name spaces are wrong because it
was before we modularised the ontology and some of the classes and
properties are deprecated so don't take it as definitive but it might
give you a few more clues.

Yours,

Faith

[sdm10012]

Thank you Leader Lawrence ...

more than enough to absorb in your reply ....

happy to evangelize your work to the New York Lotico Meetup group and would love to know about academicians / tecchies ( facebook is opening a NY engineering center ) in my area .... may be publishing my prototype soon .... for an arts education e-learning experience ...

Great Group !!1

Steve

[Faith Lawrence]

Also worth noting that you'd probably model the physical books
themselves and their publishing information with something like
FRBR-OO and then link that into the OM description (Book
is-expression-of Context) so you could use that information to infer
when order the contexts were created in by the author - even if that
doesn't tell you anything about the order that the events occur in
respective to each other, just the order of writing.

All the best with the prototype - we look forward to seeing it/hearing
all about it.

Yours,

Faith

> --
> You received this message because you are subscribed to the Google Groups
> "Digital Narrative" group.
> To view this discussion on the web, visit
> https://groups.google.com/d/msg/digital-narrative/-/tPlys9vMUmAJ.
>
> To post to this group, send an email to digital-...@googlegroups.com.
> To unsubscribe from this group, send email to
> digital-narrat...@googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/digital-narrative?hl=en-GB.

[sdm10012]

Thanks again ... actually the subject area is for the Wagner Ring Cycle operas .... 4 Operas that are the children of the Ring Cycle collection.

I took a quick look at FBRR-OO ... adds a few dimensions to the ontology .... seems like the libretto and score would be at the same level as a physical book entity  .... and then there is the performance ..... just working on the basics now ... but that elaboration may make some sense as I progress and make further associations.

Steve

> To post to this group, send an email to digital-narrative@googlegroups.com.

> To unsubscribe from this group, send email to

> digital-narrative+unsubscribe@googlegroups.com.

[sdm10012]

This issue is re-surfacing as I start dealing with Acts and Scenes in each Opera .... 

I have not run across any examples that model those types of instances .... Right now I'm "nesting" 

• Operas within the Ring Cycle "Opus" 
  
• Acts within each Opera
• Scenes with each Act
• Timelines / Occurences / Events within each Scene 

Opus, Opera, Act, Scene are all Contexts ??  Is that the correct convention ??   Seems obvious, but I have not seen examples that get that granular.

Thanks

Steve

[Faith Lawrence]

On 12 January 2013 15:58, sdm10012 <steve...@gmail.com> wrote:
> This issue is re-surfacing as I start dealing with Acts and Scenes in each
> Opera ....
>
> I have not run across any examples that model those types of instances ....
> Right now I'm "nesting"
>
> Operas within the Ring Cycle "Opus"
> Acts within each Opera
> Scenes with each Act
> Timelines / Occurences / Events within each Scene
>
> Opus, Opera, Act, Scene are all Contexts ?? Is that the correct convention
> ?? Seems obvious, but I have not seen examples that get that granular.

I'm not sure that they are all different Contexts. Think of a Context
as an area of shared truth. All the entities within that Context have
the same level of reality relative to each other.

So the Ring Cycle "Opus" probably is one Context because there are
many different versions of the Ring Cycle and it is grouping them
together in a conceptual group of shared 'Ring Cycle-ness' which would
include the different versions of the Ring Cycle.

The Opera would be a Context - because that would describe the reality
in which the narrative of the Opera takes place. Now unless the
Events/Entities described in the Acts and Scenes take place in a
different layer of reality to each other and to the general reality of
the Opera then they will just exist in the same Context. If, for
example, Siegfried had a vision of something within the course of the
narrative then the narrative of the vision would exist in another
Context - but that is unrelated to the written structures of the
Opera. The written structures of the Opera (Acts, Scenes etc) are a
facet of the way that the author chose to express the narrative within
our Context of 'Reality' and so completely removed from the reality
which exists within the world described within the Opera.

So you will probably have one Timeline to represent the linear
narrative of the Opera. On it there will be Occurrences of Events, and
the Events will exist within the Opera Context (unless the Acts/Scenes
are separate levels of reality - I don't know the Opera well enough to
definitely say they are not, but I don't think they are). Those
Events may also have a reference in them to the section of
text/audio/video which they are describing.

e.g. (Event from Midsummer Night's Dream (context/1) where Hero
(character/12) leaves a clearing in the Wood (location/3)

<rdf:Description
rdf:about="http://contextus.net/resource/midsum_night_dream/auto/event/176">
<_11:exists-in
rdf:resource="http://contextus.net/resource/midsum_night_dream/auto/context/1"/>
<_11:from rdf:resource="http://contextus.net/resource/midsum_night_dream/auto/location/3"/>
<_11:has-subject-entity
rdf:resource="http://contextus.net/resource/midsum_night_dream/auto/character/12"/>
<_11:precedes
rdf:resource="http://contextus.net/resource/midsum_night_dream/auto/event/177"/>
<_11:follows
rdf:resource="http://contextus.net/resource/midsum_night_dream/auto/event/175"/>
<rdfs:label>her. leaves</rdfs:label>
<rdf:type rdf:resource="http://purl.org/ontomedia/ext/events/travel#Travel"/>
<rdfs:seeAlso
rdf:resource="http://www.perseus.tufts.edu/hopper/xmlchunk?doc=Perseus:text:1999.03.0051:act=2:scene=2#xpointer(//div2/sp[23]/l[12]/stage)"/>
</rdf:Description>

(You can use the stuff in Media.owl to do the link if you want be
we've just been using rdfs:seeAlso until the various reference
mechanisms get sorted out and then we'll deprecate locspec and use
those instead. There's a text reference standard that seems to be
developing among the classical lit people but I can't find the link
immediately - I'll know more next week because one of the projects I'm
working on will hopefully be using it - until then there is
xpath/xpointer).

Does that make sense?

Faith

[Steve Mintz]

Yes, thank you ... makes a lot more sense re:

• getting deeper understanding of Context
  
  
• where to reference the structural elements that the author formalizes the narrative -- Acts ( and to an extent Operas) are more "artificial" than Scenes ... which relate directly to locations
  
  Many studies do describe Scenes in terms of the location and characters within the scene .... I'll look at the XPointer approach creating that type of index to the "playbill" or in the case of literature ... the Table of Contents
  
  Was not looking forward to needing to write convoluted SPARQL queries to traverse timelines / occurrences / events within nested contexts 

Being agile about this .... just putting together a couple of Scenes within the drama ,,, then code something to parse the elements of the ontology schema  and instances to embed them in attributes of an HTML5 / RdfA URL ...

Anyone else done any experimenting with that ??  

Thanks and best !

Steve

Faith

--
You received this message because you are subscribed to the Google Groups "Digital Narrative" group.

To post to this group, send an email to digital-...@googlegroups.com.
To unsubscribe from this group, send email to digital-narrat...@googlegroups.com.

[Faith]

*cough* And as Steve kindly pointed out - by Hero I actually mean Hermia (that'll teach me to actually check the RDF).

[sdm10012]

is there a conventional ontology that is a framework for describing Drama ... the Tuft's Persus project spits out XML with the proper tags, but I can't find an RDF version of their "scaffold".

Closest I could find is the BBC "Episode" graph ... but it does not have granularity beyond a "work".

There must be something !

best !

Steve

[Faith]

Sorry - I honestly thought I had replied to this.

You best bets are probably having a look at the TEI module on performance texts and FRBR (or FRBR-OO). Perseus texts are pretty much all in TEI/XML.

They've also started using Canonical Text Services (http://chs75.chs.harvard.edu/projects/diginc/techpub/cts) for references to chunks of text, something which seems to be gaining ground in the humanities and is worth a look at. I'll actually be out in the US next week learning how to set up a CTS server (and possibly seeing monster trucks?!?) so I'll be able to report back on that when I'm back in the UK (the CTS not the monster trucks unless anyone particularly cares about the trucks).

All the best,

Faith

[Paul Rissen]

Also worth saying on this that Mike and I have been recently plotting updates to the Stories Ontology, including *possibly* having a stories:World (or stories:Universe) class to capture the overall thing (a lot like context, I guess)

--

You received this message because you are subscribed to the Google Groups "Digital Narrative" group.

To unsubscribe from this group and stop receiving emails from it, send an email to digital-narrat...@googlegroups.com.

To post to this group, send an email to digital-...@googlegroups.com.

Visit this group at http://groups.google.com/group/digital-narrative?hl=en-GB.

For more options, visit https://groups.google.com/groups/opt_out.