[euopendata] Terminology: PSI-data vs Open Data

Wed Jan 12 11:50:12 UTC 2011

+1 Antti. ;-)

The idea of a clear, explicit definition of 'open data' is partly why
we started: http://www.opendefinition.org/

We're currently working on building consensus about this in relation
to 'open government data', by talking to a range of high level
stakeholders to encourage the emergence of some consensus between key,
leading initiatives in this area.

PSI is not to be conflated with 'open government data' but is a much
broader term with a longer history. PSI policy *may* include
discussion of 'open government data', but by no means is all PSI open,
or meant to be open. 'Open government data' is a subset of 'PSI' like
'open source software' is a subset of 'software'. It is nevertheless a
useful blanket term for talking about information produced by public
bodies. I understand it is (as Antti comments) in some ways quite a
EU-centric term, but has traction and is widely recognised
internationally, e.g. used by the OECD for a meeting on this area in
2006 [1].

In my mind the terminological distinction is important so that we do
not start to create the impression that 'open data' is just government
material (PSI) that is freely available online. Crucially open data
must be usable by anyone with a minimum of restriction (at most
attribution, and, possibly, integrity or sharealike requirements).

Jonathan

[1] http://www.oecd.org/document/17/0,2340,en_2649_33757_36860241_1_1_1_1,00.html

2011/1/12 Antti Poikola <antti.poikola at gmail.com>:
> Hi Peter,
>
> Congrats for the new born!
>
> The question of Open Data vs. PSI is very relevant.
>
> In Berlin last year we drafted this picture:
> http://picasaweb.google.com/antti.poikola/Open_Data#5561247540812199682
>
> Instead of PSI in the picture there is Government Data. PSI as a term
> relates to the PSI directive (EU) and in my understanding in not so much
> used outside of Europe (correct me if I'm wrong).
>
> Idea in the picture is that there exists:
> A: Government Data (including also sensitive data and data that fall under
> the data protection laws etc.)
> - Some part of the Gov. Data is allready Open -> Open Government Data
> - Some part of the Gov. Data is not Open even if it could (not sensitive
> data, for example data that can be accessed according to the Freedom of
> Information laws) -> Public Government Data
> - Some part of the Gov. Data can never be Open.
> B: Open Data (that includes Open Government Data and also open data outside
> of the Gov, for example wikipedia)
>
> It's very true that many PSI related discussions are not discussions about
> open data ( I understand by open data something along the lines of Open
> Knowledge Definition OKD ). On the contrary what comes to the pricing for
> example, many 'traditional' re-users say that data should not allways be
> free of charge, it's more important to be able to buy data with reasonable
> price and get some sort of contract with the data provider.
>
> The basic benefit categories of Open Government Data in my mind are:
> 1. Transparency and democracy
> 2. Innovation and economy
> 3. Efficiency of the government
>
> The PSI directive mostly focuses on number 2. which may well be achieved
> even if the data is not "Open" as long as there are easy ways to find data
> and pricing is not too high etc. In Finland there is ongoing legislation
> work to make data more usable "first between the governmental bodies (nr.
> 3.)" and later on "for businesses (nr 2.)" the discussion of Open Data
> related to Transparency and Democracy is not very strong (maybe because
> transparency is not considered to be a problem in Finland).
>
> Personally I allways speak for the Open Government Data, because I believe
> that in the long run that is the way to acchieve all benefits 1,2, and 3. Of
> course the Government Data can be made easier to re-use even if it for some
> reason is not be made Open. Increasing accessibility, making the re-use
> licensing clear, reducing prices, transforming it to better formats,
> providing it trough robust interfaces, increasing the data quality etc.
>
> What I am worried about is that the term "Open" is misunderstood and
> dilluted. People think that when they are building good APIs they would be
> "opening the data" even if the licencing and pricing models are not in line
> with the definition of Open Data.
>
> There is nothing wrong in building the Gov. APIs and doing all other
> possible enhancements for increasing the re-usability of PSI, but when there
> is no intension of making the Data Open then the word "Open" should not be
> used.
>
> -Antti "Jogi" Poikola
>
> On 12.1.2011 10:32, Peter Krantz wrote:
>
> Hi!
>
> I am back at work from parental leave and started digging into
> Sweden's approach to open data. Initial findings suggest that there
> are many who confuse the terms "open data" and "PSI". It mainly boils
> down to the concept of "free".
>
> The PSI directive does not require data to be free of charge (it has
> provisions to deal with transparent pricing though). Although there is
> not a universally accepted formal definition of "Open data" most
> people who are working in this area consider a price of 0 to be a
> requirement.
>
> How can we make sure more people understand these differences? The
> term "PSI" is getting a lot of traction within the government sector
> and it is easy for people to believe that thay are dealing with "open
> data" when, in fact, they may be far from it.
>
> Regards,
>
> Peter Krantz
> http://www.opengov.se
>
> _______________________________________________
> euopendata mailing list
> euopendata at lists.okfn.org
> http://lists.okfn.org/mailman/listinfo/euopendata
>
>
> --
> --------------------------------------------
> Q: Why is this email three sentences or less?
> A: http://three.sentenc.es
>
> _______________________________________________
> euopendata mailing list
> euopendata at lists.okfn.org
> http://lists.okfn.org/mailman/listinfo/euopendata
>
>

-- 
Jonathan Gray

Community Coordinator
The Open Knowledge Foundation
http://blog.okfn.org

http://twitter.com/jwyg
http://identi.ca/jwyg