Modeling with open data

8 views
Skip to first unread message

Gerry Tychon

unread,
Dec 2, 2011, 2:18:41 PM12/2/11
to OpenDataBC
Greetings ...

This is a little off topic geographically (by a province) but I
thought I would post it for interest and to illustrate that not all
use of open data has to be in the form of "apps".

The other day I was producing some land suitability ratings for the
agricultural area of Alberta and realized that all of the "inputs"
were open data.

The soil/landscape information is open data from:
http://www1.agric.gov.ab.ca/$department/deptdocs.nsf/all/sag3252?opendocument

The climate information is open data from:
http://www.ales.ualberta.ca/rr/Research/ClimateChange/ClimateDataforAlberta.aspx
(and thanks to Dr. Andreas Hamann - University of Alberta - for
facilitating the use of this data).

The models are based on the information in the open publication at:
http://sis2.agr.gc.ca/cansis/publications/manuals/1995-6/intro.html

A very quick image of the results are show at URL:

http://xspatial.com/sites/default/files/image/ALSRS_Image1.jpg

Basically, the greener and darker, the better. The white area is not
considered agricultural and has no data.

These ratings can be used for land use planning, land value
assessment, and also climate change scenarios (since climate is an
input).

Perhaps this will stimulate some other ideas on open data use.

... gerry tychon


James McKinney

unread,
Dec 2, 2011, 2:57:49 PM12/2/11
to opend...@googlegroups.com
Are these open data? I understand that they are free of charge. But I can find no license detailing important areas such as attribution, re-use, redistribution, modification, etc.

Gerry Tychon

unread,
Dec 2, 2011, 4:20:10 PM12/2/11
to opend...@googlegroups.com
Hi James ...

That is a really good point and suitable for discussion. All of these
sources were posted to the web prior to much discussion of open data
and the best way to present it -- the web is floating in this kind of
information. Maybe some "legal types" have a more educated opinion but
I have been informed that if you want to retain "rights" (e.g., in my
case software) then you need to have an upfront license that the user
agrees to othewise you are in a very difficult position trying to
enforce "rights" after the fact. Having said that, I think it is
better for terms of use to be explicit rather than implicit. Other
than a few government organizations and groups such as ourselves most
people have a poor understanding of what issues are involved with open
data.

... gerry tychon

James McKinney

unread,
Dec 2, 2011, 5:04:37 PM12/2/11
to opend...@googlegroups.com
Where did you get that information? My understanding is the opposite. You automatically get the full strength of copyright protection on your copyrightable works. Making it available on a web site does not mean you surrender all your rights. (Obviously, it will be hard to enforce having people pay for something you are offering for free, but you can still get legal protection against reverse-engineering, etc.)

David Eaves

unread,
Dec 2, 2011, 5:22:10 PM12/2/11
to opend...@googlegroups.com, opend...@googlegroups.com
Have to reinforce James viewpoint. The absence of a license means that anything that can be copyrighted on the website is copyrighted. And the rights holder has evey right to request a take down of their material.

--
www.eaves.ca
@daeaves
Sent from my iPhone

Herb Lainchbury

unread,
Dec 2, 2011, 6:03:11 PM12/2/11
to opend...@googlegroups.com
Thanks for sharing these Gerry.  It's interesting stuff.

I wouldn't call them open though.  

This is critical.  Using data based on implicit permissions, or hoping they won't get pursued for infringement is unacceptable for people to invest the kind of time it takes to build an app or a business.  This is why the legal framework is such a big deal and why I keep advising governments to work hard to get this part of open data right.  Follow the example of the US or Australia or New Zealand or Surrey or Langley or Township of Langley or Winnipeg Transit.  Fredricton has unfortunately adopted an eerily familiar license on their brand new site.   http://www.fredericton.ca/en/citygovernment/TermsOfUse.asp   


I think there is an unintended side effect of the way open data is being rolled out in some jurisdictions.  Public data that was once generally considered usable ( financials, stats, directories, registries, codes, inspection data, schedules ) are now considered not usable because they don't have explicit permissions associated with them.  When they do get an official license applied to them, their new explicit "open" status can be significantly more restrictive than the implicit understanding prior to the movement.

Also, governments often set up a new agency that identifies data citizens can use and place it in a catalogue, and anything that's not in the catalogue then has a mysterious cloud over it by contrast, even if it was formerly considered public.  Depending on how fast this new group can get through the relevant data to catalogue it, citizens could feel they have significantly less data than they had before the official process.  Hopefully this is temporary.

I personally think Australia and the US are on the right track here making data open by default.  In Australia agencies are supposed to go through a process in order to keep data closed. (There is currently no binding law to enforce this but the intent is there).   This would be relatively easy to put in place in many jurisdictions as data is already routinely assessed for privacy concerns and treated accordingly.  Determining if it should be closed would be just another check box, and since the vast majority of public data should eventually be open data, it makes sense to identify and document the much smaller set of data destined to be closed.  Everything already on the internet should just be open.

The really ironic part of all of this that there are people out there exploiting this public but not open data while folks keen on open data are patiently waiting for the "open" label to be applied.  Why a government would release data on the internet and then NOT apply an open data license to it is beyond me.  ( Here's some data - but don't use it ).

My hope is that this formerly "free" public data will eventually be explicitly liberated at some point either by being blessed and included in an official open data catalogue or by open becoming the default.

H

--
Herb Lainchbury
Dynamic Solutions Inc.
www.dynamic-solutions.com
http://twitter.com/herblainchbury

Luke Closs

unread,
Dec 2, 2011, 6:15:27 PM12/2/11
to opend...@googlegroups.com
I would be cautious about building a business or app on data that
isn't explicitly licensed. But I wouldn't be very cautious about
building a hack on such data.

If the hack starts to grow then I think it becomes the serious issue
that Herb has pointed out. And you'd really need to consider how much
you rely on it.

But I'd encourage people to hack on data they think is open, even if
it's not explicitly so. This pushes boundaries a bit and lets us act
_as if_ herb's hope of "open by default" is true today.


Thinking about this in light of Recollect, I regularly consume
non-explicitly licensed schedule and geo data from municipalities in
the form of their Garbage schedule and zones. I think this is okay,
at least okay enough for me to keep doing until it becomes an issue.
That's a personal/team decision, others will likely come to different
conclusions given different projects and datasets.

--
Dictated but not read.
Twitter: @lukec

Dan Bonab

unread,
Dec 2, 2011, 6:57:02 PM12/2/11
to opend...@googlegroups.com
I am with you on that Luke. 
the problem I have with "everything open by default" is that it is going to make custodians of data extremely protective and they would be probably less willing to publish their data.
We can't forget about the political aspects of open data in my opinion.
 
My ideal world instead would be where every piece of data on internet is accompanied by a PDDL.

cheers

Dan


From: Luke Closs <luke...@gmail.com>
To: opend...@googlegroups.com
Sent: Friday, December 2, 2011 3:15:27 PM
Subject: Re: [OpenDataBC] Modeling with open data

Luke Closs

unread,
Dec 2, 2011, 7:13:16 PM12/2/11
to opend...@googlegroups.com
Dan how do you see that problem playing out in the US? Do you have
examples of where you see that problem in places where it's already
the case?

Dan Bonab

unread,
Dec 3, 2011, 2:41:50 AM12/3/11
to opend...@googlegroups.com
That's a difficult question because you 'll only know about the data that is published not the ones that are hidden. I was talking mostly about Canada.
Political model in US is different than up here. Parties are being more scrutinized there, so they don't have the option of hiding things as much.
Here instead citizens (like open data communities) push governments to be more transparent. I'd say political parties are a little too nice to each other here:) at least when it comes to open data.
In federal level the pressure(or stick as David Eaves says!) is even less and I don't want to get into discussion about majority government and all that.
Even in BC which is supposedly the pioneer of open government and open data (and don't get me wrong lots of people there are doing a great job of opening more and more data) in Canada, they publish the travel expense report of government officials in a closed site (openinfo.gov.bc .ca). Not to mention that all expenses are in pdf file and there is no detail (itinerary breakdown) available.
In any case that's all my personal opinion and in the previous email I was just trying to emphasize the importance of licensing for open data and that's all.

Cheers

Dan
Sent: Friday, December 2, 2011 4:13:16 PM
Reply all
Reply to author
Forward
0 new messages