Hello all,
I have been following this thread for some time now, and I feel like
there the "Real problems" title would not be justified without adding
the problem of data quality to the mix.
Regardless of the format, the good old "garbage in garbage out" should
still count. That is to say that it should count more than ever,
considering the rapidly growing mountain of data we are asking for and
dealing with. Unfortunately, it is not rarely when I am seeing that we
have only increased the quantity, and not necessarily the quality.
For example, try to geocode with "roof-top" accuracy the following
addresses located in Santa Clara, California: 2102 Town & Country Lane,
2298 Capistrano Drive, 3719 Swallow Way, 3362 Pruneridge Avenue.
Thank you,
Gabriel
On 4/11/12 12:59 AM, Stefan Keller wrote:
2012/4/10 Emilie Laffray<[email protected]>:
Hello,
I agree 100% here.
The problem here is that those questions are on the semantic level. I think
that before we start thinking about file format we should think about
organizing them in a meaningful fashion in the first place. It is where most
of the start ups are working on: how to organize the data in a meaningful
fashion and make use it. This is where the extra value is currently. Once we
have agreed on what needs to be shared, only then file formats will actually
matter (for that topic of course).
I think this is a place where the "ever so successful semantic web" might be
successful.
Emilie Laffray
Agreed.
Though I wonder why only few emphasize that there is a gap between
data (format/encoding) and information. It's called database schema or
domain model! That's why I would put a system catalog table into
something like SQLite which encodes any higher level data definition
language as key/values.
By chance, just recently another "big" worldwide dataset was
published: GADM: http://www.gadm.org/. And guess which formats they
offer for download? I cite "... shapefile, ESRI geodatabase, RData,
and Google Earth kmz"... I'd say: that's just another missed chance to
semantic interoperability because of the lack of a common "Shapefile
of the future" (accompanied with some well known data definition).
Yours, S.
On 9 April 2012 23:24, Brian Russo<[email protected]> wrote:
Knowledge> Information> Data.
At the knowledge and problem solving level you don't want to bother
with data format issues (correctly so IMO), however that doesn't mean
they don't matter.
Sent from my iPhone
On Apr 9, 2012, at 11:59, Steve Coast<[email protected]> wrote:
Nobody give a flying monkey about the format jerking.
It's 2012 and we're waffling about formats. How about problems which I
actually care about or might generate actual revenue?
* What is the bar within 30 minutes journey from me with a decent IPA, a
good atmosphere and free parking?
We can't do that.
* Which of my neighbors will lend me their circular saw for $10 so I
don't have to buy one for $300?
We can't do that.
* Where should I live so that my kids have a decent school, the taxes
are good, the commute isn't too bad and the crime rates are low?
We can kind-sorta do that. Maybe.
We live in a world where we need a GIS professional to help us with the
most basic of questions. If I may, we have a Small Data problem;
http://stevecoast.com/2012/04/10/small-data/ not any format problems.
Steve
_______________________________________________
Geowanking mailing list
[email protected]
http://geowanking.org/mailman/listinfo/geowanking_geowanking.org
