|
|
| GeoCommunity Mailing List |
| |
| Mailing List Archives |
| Subject: | Re: [gislist] But what the hell can a GIS really DO? |
| Date: |
05/13/2005 01:00:00 PM |
| From: |
Neil Havermale |
|
|
Yep - the issue of joint probabilities can confound any multilayered
analysis to a point were you might only be able to sort out statistically a
small fraction of legit insight? What you think you see may to be what
really is....
-----Original Message-----
From: Stephany Filimon [mailto:stephany@imagetext.net] =
Sent: Saturday, May 14, 2005 1:16 AM
To: gislist@lists.thinkburst.com
Subject: Re: [gislist] But what the hell can a GIS really DO?
Jeremy,
There was a debate, not too long ago, about the direction in which list =
content was headed, what the list should/should not primarily focus on.
Part =
of it was about the plethora of postings on software issues, vs. more... =
engaging? theoretical? topics such as yours.
I wouldn't abandon the list just yet. I'm currently working on visual
"maps" =
of financial data and activity, just to experiment at this stage. But I
think =
this is different from your idea of GIS models - can you elaborate? Are you
talking more about combining existing models (i.e., the statistical
likelihood =
of condition A in location B happening at C time, forecast/estimate models =
already used for weather, rainfall, financial markets, etc.) with GIS layers
(adding however many variables), thus making for a "new" GIS model of sorts?
I'm on the technical side of GIS, so I have this tendency to skip ahead and =
start thinking of the "how" - what pieces would you tie in to do something =
like this, what statistical/mathematical models would you integrate? =
I think the opportunity for rich multivariate, spatial analysis is great.
I'm =
also fascinated with (and researching for my dissertation) sources of error
in =
decision making. For example, if one is using predictive models and GIS =
layers have different sources of data (weather service, windmill sensors, =
temperature sensors, etc.), how does one appropriately consider not just the
margin for error in each layer, but in combined layers? There's also the
side =
of how human beings make decisions - and where their judgments tend to be
more =
and less accurate based on their understanding of data before them. I think
your idea of GIS models may be able to take much or all of this into =
consideration.
Quoting Sonny Parafina <sonny@ionicenterprise.com>:
> You are on the wrong listserv. Many GIS-L list members believe this is =
> alternative to commercial software support. The list traffic here is =
> typically, "Which button do I push on software X to get result Y."
> =
> I would suggest that you join the geowankers listserv, where the =
> intersection of the spatial web and semantic web is the focus.
> =
> Here is the geowankers URL:
> =
> http://lists.burri.to/mailman/listinfo/geowanking
> =
> =
> =
> sonny
> =
> =
> Jeremy Stocks wrote:
> > I have been in the field of GIS nearly 14 years. I have seen every new
> craze come and go. But something exciting is slowly happening , thanks to
the
> following factors. I shall list them:
> > =
> > a.. The web for GIS is starting to come of age. Applications which are
> able to serve large quantities of data necessary for GIS are now here. =
> > b.. The rise of free web GIS such as Mapserver which allows people to
> publish their own maps on the internet as long as they have a connection. =
> > c.. The removal of selective availability on GPS receivers is allowing
> ordinary users access to precision they other wsie would never have had.
> > =
> > These changes, particularly the web and GPS are giving ordinary users
power
> to create maps they were previously never able to do. Now I am told the
> location-based industry according to the Guardian article
> http://www.guardian.co.uk/online/story/0,3605,1453293,00.html is going to
be
> worth a lot of money in the future. =
> > =
> > The apps I am seeing are impressive but are only the tip of the iceberg.
> Everyone is serving out datasets based on exisiting features such as
roads,
> railways and contours.
> > =
> > But what would be very exciting would be what I call "conceptual" maps,
> where the GIS actually does what many of us in the industry have dreamed
it
> can do, namely the ability to create new maps based upon amalgamations of
the
> existing data layers into new products which we call GIS models. This I
argue
> is the very deep heart of GIS and what enticed many of us into the
profession
> long ago.=B4These can be used to answer questions such as:
> > =
> > 1. Tonight there is a frost. Where are the likeliest danger spots for me
on
> my local roads given the predicted temperature? Can I upload aGPS location
> back to the map server to confirm or deny the frost location, thus
improving
|
|
Sponsored by:
For information
regarding
advertising rates
Click Here!
|
