W3C home > Mailing lists > Public > public-multilingualweb-lt@w3.org > January 2013

Re: issue-68 from an annotation representation point of view, with potential implications for annotatorsRef and standoff markup

From: Felix Sasaki <fsasaki@w3.org>
Date: Mon, 28 Jan 2013 08:11:27 +0100
Message-ID: <5106249F.3020709@w3.org>
To: Yves Savourel <ysavourel@enlaso.com>
CC: public-multilingualweb-lt@w3.org
Hi Yves, all,

Am 28.01.13 00:12, schrieb Yves Savourel:
> Hi Felix, Jörg, all,
> A few notes with the proposed mechanism:
>> Comparison to current ITS1 "Terminology":
>> its-tan-type="term" plays the role of term="yes".
>> its-tan-info-ref plays the role of termInfoRef.
>> its-tan-ident-ref links to a term data base.
>> its-tan-confidence provide confidence information.
> a) We don't have a term='no' anymore. I know it's a little used feature, but the bottom line is: it goes away (you can't override a global rule that marks a given element as 'term' anymore).

Good point.

> b) Links to term DB have been put in termInfoRef so far, so we would basically split the termInfoRef features into two. I'm not against, I'm just pointing out that conversion from ITS1 to ITS2 won't be automatic.

Not sure if I understand: if we set "tan-type" to "term" and have the 
"ident-ref" attribute: wouldn't that be teh lik to the termDB?

>> - If there is inline annotation, check if there is an id attribute
>> (in HTML) or xml:id (if XML serizalization of HTML is used and
>> with lower precedence compared to id). For formats other than HTML,
>> add xml:id if possible or use the id attribute appropriate
>> for that format.
> It seems we would have two very different ways to use standoff markup: LQI and Provenance use a reference to the standoff enclosing element, here we would have the reference in the enclosing element to the local inline content. It would probably be better to be consistent.

Not sure - if the term+disambig standoff is adopted, I actually would 
rather call it different, e.g. "multilayer annotation" - and keep 
everything else as is. The rational is that the other mechanisms seem to 
work fine. So I wouldn't make the multilayer approach for disambig+term 
a big dial and new architectural principle, to be deployed for other 
data categories - but rather a means (to be defined in the respective 
sub section) trying to resolve a last call comment.

> In addition using some kind of tan-ref attribute would allow us to use a IRI and point outside of the document if needed (see Jörg's comments bout that).

Good point.

> Also, for LQI and Provenance we also have either local attributes or standoff markup, not both at the same time like in this proposal. It would be also probably better to be consistent.

See above - I think the proposed algorithm for fetching the multilayer 
info for disambig+term described in the previous mail (fetch for each 
node the annotation wrapper and resolve the IDs) is fairly simple - and 
we wouldn't need to change other parts of the spec IMO.



> cheers,
> -yves
Received on Monday, 28 January 2013 07:39:30 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 16:32:00 UTC