Anything about Tenor? & creating a content system -> RDF
Leo Sauermann
leo at gnowsis.com
Thu Aug 11 17:51:13 CEST 2005
Hi all.
>For the record there's already an ioslave playing with the C++ port of Lucene:
>http://www.kde-apps.org/content/show.php?content=23874
>http://kioclucene.objectis.net/documentation/demos/demo1
>
>
great
>Ah-ah RDFS/OWL detected. =)
>Yes that's exactly my biggest worry with RDF, it could quickly become overkill
>we surely don't want to support most of its vocabularies.
>
>
rdfs:domain & rdfs:range are usually enough. rdfs:subclassof is also
nice. but more you often don't need
>You state the obvious, the point is more do we really want and need to express
>anything?
>
>
sooner or later, its better to be on the safe side. just a hacker feeling.
>
>
>>there are some points in RDF that suck, like sequences or lists, but
>>these can be handled.
>>
>>
>
>Could you be more precise here?
>
>
http://www.w3.org/TR/rdf-primer/#containers
this gets ugly when you really add thousands of items in a container and
then add/remove a few in between. especially with MySQL backed models,
you might have surprises
nearly the same with those:
http://www.w3.org/TR/rdf-primer/
another big big and structural problem is that when you want a container
as a property of a resource, you can't tell the value type.
example: I am a folder and I contain Images, a list of images
I - foo:contain - SequenceX
I - rdf:type - foo:ImageFolder
SequenceX - rdf:li - image1
SequenceX - rdf:li - image2
If you want to express this data structure in RDFS, you can't really,
because it only allows you to either say:
foo:ImageFolder rdf:type rdfs:Class
foo:contain a rdf:Property;
rdfs:domain foo:ImageFolder;
rdfs:range rdfs:Seq.
saying, that you can't do now:
foo:contain rdfs:range foo:Image
but that is details.
>
>
>>>In this case, I'd like to be able to have a textual format in order to
>>>have a clearer opinion about RDF usage in Aduna Autofocus. I can't find
>>>such an export unfortunately.
>>>
>>>
>>ok, I'll bug the developers for this one.
>>
>>
>
>Thanks a lot.
>
>
They said they work on a developers release. Acutally you could get your
fingers on Aduna Metadata Server, thats basically embedded in
AdunaAutofocus.
>
>
>>sure, but you will need a metadata format anyhow.
>>The big ontology work is not sitting down and writing an OWL file, the
>>problem is to agree with people like the nifty guys in this email-group
>>if we now call the thing "title" or "name" or "label". That is really
>>the hard work.
>>
>>
>
>Oh please, don't get me on this... your view is only about one flavor of
>ontologies. In the general case you know that it really raises more questions
>on a structural level and the ontology refinement is a work by itself. In
>particular when you're working on domain ontologies where it's difficult to
>reach a consensus on the domain representation it's not only about naming
>some properties but which concepts you'll keep in the final version. You have
>also to be sure that it meets the requirements for the resulting application.
>
>
yep, I only illustrated the beginnning of the endless discussions. But
thats everywhere with code.
>I guess that for data formats "ontologies" are really easier to obtain, since
>the available metadata were fixed first anyway. Nice point, I didn't noticed
>it earlier, it's maybe because I have to deal with domain ontologies on a
>daily basis.
>
>
which domains?
>Yes, but we have guarantees that it's a better solution than C in our case. Do
>we have guarantees that going for RDF formalism is better than an ad hoc
>model in our case? not sure...
>
>
never sure, its just strong developer feeling.
>
>
>You should really avoid using such absolute terms. A particular community as a
>whole is moving towards RDF, it's not just because you are in this particular
>community that the world as a whole is moving towards RDF.
>
>
ok agreed. I am too deep in my topic.
>>* that metadata extractors/file filters from aduna, gnowsis, kowari,
>>intellidimension, ...
>>
>>
>
>I admit that at least Aduna looks interesting in this matter. As for gnowsis,
>it's on my todo on things I should test, but I doubt I'll find much time to
>do it for a while.
>
>
>
don't touch it until december, its dirty. we are working on it night & day.
>
>Hehe, I know another one in this regard : the KDE community.
>
>
ok, thats a point now. yep, this community is good.
cheers
Leo
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.kde.org/pipermail/klink/attachments/20050811/aa70f19b/attachment-0001.html
More information about the Klink
mailing list