Re: Collecting Best Practise Process for GLD - the 10 steps to GLD / LOD

Dear Denise and Biplav, all

many thanks for your feedback and support on this - I also do think 
that the GLD Best Practise approach is a very good one - although it is very 
detailled and technically driven - speaking with civil servants here in Austria showed
that they are looking for a more 'meta level approach' first - so we discussed
the following steps in a workshop here that could be helpful for my given example of
the Austrian city with 150k inhabitants that has already ~50 datasets as open data in a 
catalogue (in several formats) that is thinking about establishing a 'LOD based digital data infrastructure'

-----
1) Set Up a (Pilot)Project on Linked Open Data 
2) Identify basic data sets that provide most value / benefit (basic data sets could be: geo (postcodes), organisations, types of schools, code lists et al)
3) Evaluate what datasets could be linked to each other for most benefit (internal & external linking sources)
4) Clarify license issues
5) URI Konzept erarbeiten
6) Evaluate (RDF) vocabularies and specify the fitting vocabs for your data
7) Establish an infrastructure (Triple Store, Endpoint, web API on top of LOD, ...) 
8) Triplification: convert the spefied data sets to RDF
9) Link the datasets / the data (internally & externally) 
10) Publish the data and spread the word about this
11) Continuous data maintenance and add. data publication
12) Establish & manage network effects with other data holders
13) Support the use of the data / of the data infrastructure - if possible provide funding for 1st projects (internally & externally)
    E.g. funding, VC start up sessions (competitions) - but also support in LOD principles & technologies, events, manuals, etc...
-----

Feedback very welcome - regards - martin

----- Original Message -----
From: "Denise Warzel (NIH/NCI) [E]" <warzeld@mail.nih.gov>
To: "Biplav Srivastava" <sbiplav@in.ibm.com>
Cc: "Martin Kaltenböck" <m.kaltenboeck@semantic-web.at>, public-gld-wg@w3.org
Sent: Wednesday, 12 June, 2013 3:47:34 PM
Subject: Re: Collecting Best Practise Process for GLD - the 10 steps to GLD /  LOD

Very interesting question. I think the best practices w3c published are very good. I'd emphasize the modeling aspect and the use of standard vocabularies ... Butor just for the publishing of the statements ( predicates) but Alison for the model....The data that is to be linked should  needs to share some aspect of their mostly, and those should have the same Uris... Pre-planning around harmonized models should help.... Otherwise you will have LOD data sets that are silos, not sharing common subjects or objects- even if they share common predicates....

Denise with help from my iPhone

On Jun 11, 2013, at 11:33 PM, "Biplav Srivastava" <sbiplav@in.ibm.com> wrote:

> Hi Martin,
> 
> In the draft best practices note [1], please see Section 1 - summary. It
> lists down 14 items. Will they suffice? A lot of members have given
> comments to arrive at them.
> 
> My own experience is that the following factors are very important in
> helping a city open up their data:
> 
> a) Familiarity with regulations regarding data sharing. The city
> authorities may need to be educated about what can and cannot be opened.
> The regulations differ at countries, state and even city levels. The aim
> should be to share as much as possible without limiting it by the current
> thinking of usage possible. Typically, any outcome of an initiative with
> publicly funding can be shared. Anything which is personally identifiable
> must be masked before sharing.
> b) Having a process and tool to create data. The city's effort will be as
> good as the volume and quality of data it actually makes public. They are
> worried if they will make a mistake and cannot retract if things go wrong.
> Their efforts are more likely to succeed if they can use a well-tested
> process or set of tools that they know has worked for someone, and can thus
> address common issues (versioning, URI, ...). That is why, publishing
> successful case studies is so important.
> 
> The work group decided not to prescribe any specific set of tools.
> 
> 
> [1] https://dvcs.w3.org/hg/gld/raw-file/default/bp/index.html
> 
> 
> 
> Regards,
> --Biplav
> 
> 
> 
> 
> 
> 
> 
> 
>  From:       Martin Kaltenböck <m.kaltenboeck@semantic-web.at>                                                                    
> 
>  To:         "public-gld-wg@w3.org" <public-gld-wg@w3.org>                                                                        
> 
>  Date:       10/06/2013 01:49 AM                                                                                                  
> 
>  Subject:    Re: Collecting Best Practise Process for GLD - the 10 steps to GLD /  LOD                                            
> 
> 
> 
> 
> 
> 
> Dear all
> 
> a short addendum to my email below - I think I forgot to mention an
> important point - that I am looking for:
> 
> The '10 steps' to LOD to become the digital infrastructure for a
> governmental department / city , ...
> 
> So what would be the most important steps to build such a data
> infrastructure using LOD principles et al for city X with 150k inhabitants?
> 
> Looking forward to your feedback, best! Martin
> 
> Martin Kaltenböck, CMC
> Managing Partner, CFO
> Semantic Web Company
> 
> Web: http://semantic-web.at
> Blog: http://blog.semantic-web.at
> Meta Data Management: http://poolparty.biz
> Phone: +43 1 402 12 35 - 25
> 
> 
> 
> Am 09.06.2013 um 17:39 schrieb Martin Kaltenböck
> <m.kaltenboeck@semantic-web.at>:
> 
>> Dear all
>> 
>> I had interesting discussions last week with Austrian public servants
> about 'best practise for a process / for a linked open data strategy in
> public administration'
>> and thereby my question to all of you comes up if there is some
> information available in the GLD working group and / or any other feedback
> by
>> the group or tipps or ideas on that....many thanks for feedback on this
> (see more below).
>> 
>> Fore sure this is also around best practise of 'publishing linked open
> data' (but for this issue lots of good best practise infos seem to be
> available)
>> 
>> BUT: the question is more around:
>> 
>> When a department of public administration starts to publish linked open
> data (a department, a city, a region or the national gov -
>> no matter if open data is already in place or not):
>> 
>> - what are the most important issues to keep in mind (license, schema
> selection, technologies, URI concept, ....)?
>> - what are the most important known bottelnecks and / or pitfalls?
>> - what is important by selecting the respective data sets (is it
> important to plan around what data to publish first) - e.g.: publish 'basic
> data sets'
>> first like spatial data, organisations as LOD et al - so that data sets
> that are published later on can be linked to this basic data sets
>> - what comes next - after the 'pure publishing'
>> - what else comes to your mind...?
>> 
>> Remark: Imagine a small city with 150k inhabitants invites you to give
> them guidance / help for their LOD strategy - asks you for 'the 10 most
> important
>> steps to LOD' for their city - what would you tell them?
>> 
>> Output of this collection could be a 'best practise guide / 10 steps to
> LOD in public administration'...maybe something like this already
>> exists - but I am not aware of it - so many thanks for feedback - any
> ideas are very welcome - best regards from Vienna - martin
>> 
>> 
>> --
>> Martin Kaltenböck, CMC
>> Managing Partner, CFO
>> 
>> Semantic Web Company (SWC)
>> Mariahilfer Strasse 70 / 8
>> A - 1070 Vienna, Austria
>> Tel +43 1 402 12 35 - 25
>> Fax +43 1 402 12 35 - 22
>> Mobile +43 650 3905697
>> 
>> http://www.semantic-web.at
>> http://blog.semantic-web.at
>> http://poolparty.biz
>> 
>> LOD2 - Creating Knowledge out of Interlinked Data - http://lod2.eu/
>> OKFN-AT - http://okfn.at
> 
> 
> 
>

Received on Thursday, 13 June 2013 13:22:42 UTC