On 6/8/11 11:13 AM, Christopher Gutteridge wrote:
All in all, if we can write a library which can read a page with
schema.org data encoded and spit out RDF, who cares?
Exactly!!
I'm looking the examples, and it looks easy enough to turn into
triples, albeit there will often be graphs with nothing but bnodes.
There's no way normal web developers will assign URIs to things until
they see the benefit... Could we suggest a trivial extension to
schema.org to let people add URIs for itemtypes.
<div itemscope itemtype=*MailScanner has detected a possible fraud
attempt from "schema.org" claiming to be*
"http://schema.org/Organization"; about=*MailScanner has detected a
possible fraud attempt from "totl.net" claiming to be*
"http://totl.net/#org";>
If I had more hack-slack I'd knock up a library which takes a
schema.org encoded page and spits out triples.
Are people going to create some semantic abominations using
schema.org? of course, but they were already able to do that in RDFa.
Yep! Ditto RDF. You can make crap using any markup :-)
This is going to be used by the same kind of people who were
implementing RSS a decade ago. Just accept that the world is going to
end up with a big pile of shonky data!
But therein also lies opportunity. In the world of business only one
thing rules, and its called: opportunity cost. Commercial entities track
opportunity costs like chancellors and finance ministers track
inflation. When someone else is eating their lunch they adjust. Google,
Microsoft, Yahoo! haven't decided to push structured data via graphs via
HTML out of the goodness of their hearts, they're already incurring
opportunity costs which they aren't stupid enough to announce to the
whole world.
schema.org is so very much more human-readable than RDFa. It wins
hands down on that.
Microdata for the audience schema.org is aimed at trumps RDFa, big time!
But at the end of the day, who cares about syntaxes for graphical data
representation, we should all be happy about the fact that graphical
data representations are getting broader acceptance.
Transforming graphical structured data (in any form) to pure functional
Linked Data is an exponentially simpler job than starting with some
totally unknown from of structured data or binary blobs.
We're the linked data community. RDF is a tool to an end, no more.
+1000
Rather than sit around and feel glum about this coming wave of data
being a bit wrong, we should be looking at how to help it become
Linked (and Open).
+1000
I think we made a big mistake in using http URIs. It's too confusing.
Like RDF, HTTP URIs are an option.
Personally, the problem lies in a narrative that refuses to dissociate
RDF and the concept of Linked Data Structures embellished with URIs.
There are too many distracting wars about superficial matters while the
core concept of leveraging URIs for Object Naming, Object Addressing,
and Object Representation is totally lost. We oscillate between URI and
URL which solves zilch since the URI abstraction serves two needs: Name
and Address, which are put to max use via Linked Data concept.
If we'd used <thing://totl.net/> with the convention that you can find
facts about it by converting "thing:" to "http:" then the world would
be much less confused about URIs.
I don't think so.
The world (beyond RDF and Semweb) actually understands de-reference
(indirection) and address-of operations. They worked with these
mechanics in other realms esp. pre WWW. The trouble with RDF and Semweb
narratives is that they are pretty poor when it comes to making
connections with what already exists and what people may already understand.
There are more programmers (generally) that understand what I refer to
above than there are folks that actually cling to RDF and its usage for
graphical data representation as some new phenomena in computing. This
is why I continue to state: RDF narrative is broken.
I know this is probably an old topic of conversation, but it's a
massive impediment to the public understanding of URIs for things not
available via the HTTP protocol.
Yes!
HTTP is a cost-effective option for Object Names due to WWW ubiquity,
but that doesn't make it the sole option. For reasons I'll never
understand, we are now in an era of boolean "OR". Basically, if I am not
the sole option I am a failure! Come on! Is the WWW about sole options
or the ability to expand endlessly via hyperlinks? The loosely coupling
core of AWWW is what makes the WWW work. The fact that HTML worked for
Documents doesn't mean the W3C has to prescribe RDF for data. Standards
aren't prescribed, they become what they are after significant adoption.
RDF doesn't have any kind of significant adoption relative to the WWW to
which its being forcefully prescribed.
I'm amazed that people are so surprised about schema.org. Don't you
realise that RDFa, RDF/XML and using http:// URIs for real world
things is really really confusing and make the amazingly useful idea
of Linked Open Data much harder to get to groups with?
+1000
These days when I teach people about RDF data I start with N-Triples
as it's the easiest format to grok.
I've never taught or explained the concept of Linked Data via any
markup. I start by explaining 3-tuples, de-reference (indirection),
address-of etc.. I focus on showing their existence in languages, apps,
and middleware that preceded the WWW.
Sorry for getting a bit ranty, but this community has no focus on
lowering the barriers which make it hard for the mainstream web
community to start producing linked data. I find that very very
frustrating.
s/no/to :-)
+1000
Kingsley
Harry Halpin wrote:
On Sat, Jun 4, 2011 at 2:51 PM, Michael Hausenblas
<[email protected]> wrote:
All,
Thanks a lot for the comments we received so far, both here and (even more)
off-list. Now, to make our life a bit easier, may I ask you to provide
suggestions concerning the mapping (or feature requests alike) directly to
the Github [1]? Of course, if you're more into it, feel free to clone the
repo and issue a pull request.
As you can imagine, this is a community endeavour - we just happened to kick
it off ;)
Actually, I'm also going to point out that the W3C asked for EU
funding about a year ago for something *very* similar - and at the
time had the interest even of Google - for hosting a RDF version of
something quite similar to schema.org. But thanks to the wonderful
judgement of the reviewers of EU proposals and the Semantic Web
academic community, we weren't given funding :)
Obviously Google and friends see a good opportunity here to actually
make a place to find structured data vocabularies on the Web. While I
wish they had better support for RDFa, I can see that the whole
RDFa/microdata/microformats lack of convergence is causing a confusing
mess for ordinary webmasters.
cheers,
harry
Cheers,
Michael
[1]https://github.com/mhausenblas/schema-org-rdf/issues
--
Dr. Michael Hausenblas, Research Fellow
LiDRC - Linked Data Research Centre
DERI - Digital Enterprise Research Institute
NUIG - National University of Ireland, Galway
Ireland, Europe
Tel. +353 91 495730
http://linkeddata.deri.ie/
http://sw-app.org/about.html
On 3 Jun 2011, at 22:06, Michael Hausenblas wrote:
http://schema.rdfs.org
... is now available - we're sorry for the delay ;)
Cheers,
Michael
--
Dr. Michael Hausenblas, Research Fellow
LiDRC - Linked Data Research Centre
DERI - Digital Enterprise Research Institute
NUIG - National University of Ireland, Galway
Ireland, Europe
Tel. +353 91 495730
http://linkeddata.deri.ie/
http://sw-app.org/about.html
--
Christopher Gutteridge --http://id.ecs.soton.ac.uk/person/1248
You should read the ECS Web Team blog:http://blogs.ecs.soton.ac.uk/webteam/
--
Regards,
Kingsley Idehen
President& CEO
OpenLink Software
Web: http://www.openlinksw.com
Weblog: http://www.openlinksw.com/blog/~kidehen
Twitter/Identi.ca: kidehen
