Intro and objectives

Hello

I'm Dan Brickley, AC Rep for Vrije Universiteit Amsterdam, where I
work on the NoTube project, which is about the collision of Web and
TV, see http://blog.notu.be for more on that. I've been involved in
W3C since 1997 when I showed up as a digital libraries-ish guy, and
got involved in the early W3C metadata project which eventually gave
us RDF, the Semantic Web and Linked Data. For a few years I worked on
the W3C team directing the SWAD-Europe project that helped move RDF
things along (supporting RDFCore WG, SKOS etc.), and at same time
started the informal FOAF project, which is an experiment in using
linked RDF documents to described linked real world entities.

None of which is particularly geo-centric, and maybe that explains my
biases here, so I'll try to make them explicit.

Some working assumptions -

* the big challenge with geo-related data isn't the "geo"-bit, but the
"related" bit. We have mountains of geo standards already, at every
level of detail. And a lot of data, increasingly public.
* the big challenge with standards is knowing when to stop
standardising, and how to let other people's work drop in to fill the
gap.
*  whatever we do here will sometimes get used in XML contexts,
sometimes in HTML, sometimes in JSON; we will need some abstraction
across notations, without getting lost in abtraction for its own sake.
* 'augmented reality' isn't a technology as such, but more a slogan
and an inspiration for those of us excited by the idea that
information can be superimposed on everyday life in a variety of ways;
visually, audibly, or even on good old fashioned maps. We will do
things to advance AR, and advance the Web, without having to define
'AR" formally.

On the geo front, I do edit the W3C SWIG RDF 'basic geo' vocabulary,
and was co-organizer of the recent Augmented Reality Developer Camp
here in Amsterdam. But I am more a Web of Data person than a
geo/maps/mobile one. I have a lot of friends in the terribly named
'geowankers' list,
http://geowanking.org/mailman/listinfo/geowanking_geowanking.org and
do my best to keep up with trends. I spend most of my standards time
these days around 'social web' standards, including co-co-chairing the
W3C SocialWeb incubator group, which is just now wrapping up.

I'm happy that this group is beginning from a very basic notion, that
of providing info about 'mere' POIs; this shows respect for the other
large efforts already well underway (GML etc), and a modesty that
means we might actually achieve something in the next few months. In
the pre-WG period I tried to emphasise the difficulty of even defining
a POI, by giving an off-the-top-of-my-head list of potential POI
entities, http://lists.w3.org/Archives/Public/public-poiwg/2010Aug/0031.html
and asking which ones are in scope. Revisiting that, I'm left quite
certain that our most important duty here is to define extensibility
and linkage that connects very basic bits of location data with
arbitrary rich information about the entity that is being located;
from DVDs in a shop to historical figures associated with a gazateer
entry. If we do that, we've done something new, something valuable,
and something that builds on the works of others rather than
re-treading old ground.

I should also mention some related work I'm interested to connect to this WG -

In 2003, several of us in the W3C Semantic Web (RDF) Interest Group
made a very basic schema, essentially lat/long/alt, and bound to
WGS84. This is very minimalist but has found some use within the RDF
datasets out there. If an outcome of these discussions was a proposal
to do anything in that space (including updates/edits/fixes to the
basic schema), I'd be happy to help. http://www.w3.org/2003/01/geo/

All of the above is pretty much my personal perspective. I have many
colleagues working in the W3C standards scene, mostly around cultural
heritage, semantic web, SKOS etc, but a few with much more geo
expertise and background than I have. I'll talk to them and maybe
they'll participate too, otherwise I'll try to proxy for them too.

So, to recap, I hope we can emphasise ways of linking basic POI
descriptions to full machine-friendly info about the POI entity,
regardless of whether it's something to buy, somewhere to eat,
something that happened or something that hasn't happened yet but
might. Such an extensibility framework should use the core technology
of the Web (links between URI-identified things) to offload work from
this group to the rest of the world. If we manage this, we take
pressure off of gazetteers and mappers to say everything about
everything, since filters, searches and other info navigation tasks
will be able to exploit external linked data as well as the core POI
descriptions.

cheers,

Dan

Received on Tuesday, 26 October 2010 12:17:07 UTC