W3C home > Mailing lists > Public > public-vocabs@w3.org > June 2014

Re: schema.org markup for a translated CreativeWork?

From: Peter F. Patel-Schneider <pfpschneider@gmail.com>
Date: Fri, 27 Jun 2014 08:24:23 -0700
Message-ID: <53AD8CA7.9030209@gmail.com>
To: Peter Lejeck <peter.lejeck@gmail.com>, public-vocabs@w3.org

That's an interesting task you have, particularly as Anime is likely to be one 
of the more difficult kinds of TV creative works to set up.  However, the 
problem of variant works pops up a lot in movies, where there may be preview 
versions, festival versions, an original release version, a director's cut, 
dubbed versions, subtitled versions, edited versions, and censored versions, 
let alone versions created later, such as colourized versions :-(, anniversary 
versions, and so on.  Most of the above attributes can be combined, and with 
current technology a single video track can be combined with multiple audio 
tracks and subtitles.  All this complexity may not matter when you are asking 
"Who starred in Gone With the Wind?", but does matter in lots of cases, for 
example if you are trying to figure out how long the movie you want to watch 
is or who is speaking at a particular time.

So do you want to do something that is going to work in the simple cases (and 
maybe fail badly in other cases)?  If so, go with your second option, even 
though the markup is confusing as to what the meaning of the language property 
is.  (Is it the language that the performer was speaking during his/her/its 
performance?  Is it the language that the performer was speaking on the 
original release version?  Or is it the something to tie the performance to a 
particular version of the work, as you intend.)  If you need something that 
can handle the non-simple cases, however, as you say, you will need to have 
multiple versions of video creative works and some way to relate them 
together.  Absent some much more powerful representation language, I don't see 
that anything significantly different is reasonable.

Whether either of these approaches fit into schema.org is a separate question. 
  I'm quite sure that in either approach you will need to use properties that 
are not in schema.org.

Anyway, good luck!

Peter F. Patel-Schneider

PS:  Some of the above is the result of a significant attempt to provide a 
conversational interface to an old system that takes an extreme version of the 
versioning approach.

PPS:  I noticed that hummingbird.me does not appear to be usable at all 
without logging in.  Boo!

On 06/25/2014 08:11 PM, Peter Lejeck wrote:
> Hi,
> I'm a programmer working on Hummingbird.me, and I'm trying to mark up
> our encyclopedia pages for anime series with microdata.  However, I'm
> not sure how I should be marking up the actors by their language.
> There's basically two approaches I can think of, but neither seems
> possible within the current properties:
> 1. Create multiple TVSeries for different languages, and put the actors
> in each.  This is unambiguous in the odd cases where there are multiple
> translations to a single language (I can think of at least one instance
> in our database where this happened).  The inLanguage property already
> exists to represent the language of a CreativeWork, I'd just need a way
> to mark them as related works.
> 2. Add a language property to PerformanceRole.  This would obviously be
> much simpler markup, but it would also add ambiguity to some cases.
> Am I overlooking something, or do you have suggestions for how to do
> this?
> Thanks,
> Peter Lejeck
Received on Friday, 27 June 2014 15:24:55 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 17:49:32 UTC