ProjectDataAgenda20100217

Dbennett — Wed, 10 Feb 2010 18:58:06 GMT

Dbennett: New page: <h1>Draft Agenda for February 17th</h1> By teleconference and IRC * Introductions (5 min) * Discussion of Scope and Interests of Participants(10 min) * Discussion of Statement of Work f...

<h1>Draft Agenda for February 17th</h1>

By teleconference and IRC

* Introductions (5 min)
* Discussion of Scope and Interests of Participants(10 min)
* Discussion of Statement of Work for this Project (10)
* Identifying Resources and Project Management for this Project (10)
* Work on First Project Step (15 min)

== How-to join - Phone ==
;eGov IG teleconferences information:
: ''Day:'' Wednesdays (every other week)
: ''Time:'' 9:00-10:30 EST ([http://www.timeanddate.com/worldclock/ lookup other timezones])
: For project meeting times, see [[Projects]]. Other information is the same.
: ''Phone number:'' +1.617.761.6200 (Boston) or +33.4.89.06.34.99 (Nice) or +44.117.370.6152 (Bristol)
: ''Conference code:'' 3468 ("EGOV")

Please, remember to dial "#" right after the conference code. If everything goes well, you should hear a beep meaning you've just joined the conference, and could hear other participants.

The teleconference system at W3C is called Zakim. There are more [http://www.w3.org/2002/01/UsingZakim detailed instructions] on how to use it. Zakim has an accompanying and useful IRC bot. Please read next section to learn more about it.

;eGov IG Teleconference IRC information:
: ''Server:'' irc.w3.org
: ''Port:'' 6665 (try port 80 if you have trouble)
: ''Channel:'' #egov

Dbennett — Wed, 10 Feb 2010 13:34:35 GMT

Dbennett: /* Data Management */

'''DRAFTY, BUT GOOD ENOUGH. PLEASE SIGNUP USING [http://www.w3.org/2002/09/wbs/33280/project-signup/ Web Poll]'''

= About This Document =

== Project Template ==

For each project, there should be a table (as below) briefly answering certain key questions:

;Community
: (Who are the people directly affected by this project? Could also be called "target audience" or "users".)
;Problem
: (What problem is the community experiencing that we can help solve?)
;Goal
: (What will things be like when the problem is solved?)
;Solution Scope
: (What kinds of solutions will we be attempting or avoiding?)
;Participants
: (Who is willing to put in time to solving this problem? WikiNames are good here, and maybe put the number of hours per week you can commit in parens after your name)
;Status
: (What is currently going on with this project? What phase is it in?)

After this table, there is room for general discussion about the project.

== Terminology ==

'''OGD''' is Open Government Data, data that governments have that they are making available in machine-readable form (more than paper or PDFs) to the public. OGD typically uses XML, CSV, custom data formats, and/or Web APIs.

'''GLD''' is Government Linked Data, which is OGD that's using the Semantic Web standards (eg RDF), and also following the Linked Data architecture, where items are identified by URIs which lead to information about the items.

-----

= W3C eGov Projects (PROPOSED) =

== GLD Tech ==

;Community
:
;Problem
:
;Goal
:
;Solution Scope
: Create and publish an W3C Interest Group Note:
;Participants
: [[User:Bandholtz|Thomas Bandholtz]] (3) - see my posting [http://lists.w3.org/Archives/Public/public-egov-ig/2009Nov/0045.html] [[User:ChrisBeer|Chris Beer]] - happy to translate into "laymans terms" as well as add what I can
:
;Status
:Early conceptualization

'''Government Linked Data: Techniques and Technologies'''

Document description: Written for someone with a basic technical knowledge of deploying web servers, relational DBMS's, client-server architecture, etc. Assumes the reader is already convinced of the merits of LD.

Some agenda topics:
* how does linked data support (non-RDF) data consumers?
* strategies for modelling government data
* essential metadata for Government Linked Open Data (eg VoiD)
* expressing rights and licensing information
* approaches to provenance, authority and trust
* using RDF for Statistical Data

== GLD Strategy ==

;Community
:
;Problem
:
;Goal
:
;Solution Scope
: Create and publish an W3C Interest Group Note
;Participants
: [[User:Hughb|Hughb]] (1) - doing this one for the learning
: [[User:Bandholtz|Thomas Bandholtz]] (3) - some involvement in Germany and on the European level
: [http://ambur.net/bio.htm Owen Ambur] - My participation is conditional: If a set of measurable goals and objectives are outlined, I will render them in StratML format.
: [[User:SWebb2|Sylvia Webb]] (4) - Interested from ANSI ASC X12 perspective
;Status
:Early conceptualization

'''Government Linked Data: Strategies and Success Stories'''

Document description: A high-level planning/roadmap document, motivating the adoption of linked data technologies for government open data, and detailing the organizational steps to a successful deployment. Written for middle and high level IT managers. Does not assume the reader is convinced about using LD.

== Web Tech ==

;Community
: Staff of government entities looking implement or improve government services with Web technologies.
;Problem
:
;Goal
:
;Solution Scope
: Create and publish an W3C Interest Group Note
;Participants
: [[User:Hughb|Hughb]] (1) [[User:Briangryth|Briangryth]] (1) [[User:ChrisBeer|Chris Beer]] (5)
;Status
:Early conceptualization

'''Best Practices for Using Web Technologies to Deliver Government Services'''

Document description: Collection of everything we know about eGov, but not going into a lot of depth. Maybe a few technology details, but still usable for non-techies. The document will address both "Web 1.0" and "Web 2.0" technologies and how these technologies can help improve government services. The document will be useful to both IT staff and policy-makers and will direct the readers, when possible, to other materials for more in-depth information.

== Data Management ==

;Community
:
;Problem
:
;Goal
:
;Solution Scope
: Create and publish an W3C Interest Group Note
;Participants
: [[User:Hughb|Hughb]] (1)
: [[User:ChrisBeer|Chris Beer]] (2-3)
: [http://ambur.net/bio.htm Owen Ambur] - My participation is conditional, based upon the willingness of the group to address the need for a standard, minimum set of metadata elements in machine-readable (XML) format.
: [[User:SWebb2|Sylvia Webb]] (2-4)

;Status
:Early conceptualization

'''Best Practices for Government Data Management'''

Document description: Focusing on the standardization of data and metadata to better help with:
* Access to the data for humans and machine/software
* Referencing the data and repositories of new or legacy data
* Picking metadata standards for embedding within or outside of the data
* Standards for aggregating or cataloging data
* Entity (proper noun) naming standards
* Considering permanence and provenance of data
* Connecting the paper/physical data or objects to the digital forms of data (bar coding, etc)

== Social Media ==

;Community
: Government employees, at all levels, who would like to use social media tools and services.
;Problem
: Justifying the use of such services in light of the security risks and records management problems associated with them
;Goal
: To collaborate with other organizations and groups to provide and outline the best practices for using such services to advance the mission of government agencies and engage the public. The best will also highlight how security risks can be mitigated and records management requirements can be met.
;Solution Scope
: Create and publish an W3C Interest Group Note
;Participants
:[[User:Briangryth|Briangryth]] (1-3) [[User:ChrisBeer|Chris Beer]] (5-10) [http://ambur.net/bio.htm Owen Ambur]
;Status
:Early conceptualization

'''Issues and Best Practices in Government Use of Social Media'''

Document description: Propriety, ethics, how officials ought to engage, .... everyone else is already talking about it already, but let's try to bring what we can together with our international, standards-based perspective. Usage of such services should be directly linked to (aligned with) objectives documented in agency strategic plans and employee performance plans. The document should also address:

* How social media advances the mission of government?
* How social media usage can increases efficiency and productivity?
* How social media tools can support the implementation OGD and GLD?
* How social media tools can interfaces and support current government Web presences?

== GLD Demo ==

;Community
: People trying to decide how much to support Government Linked Data (GLD) efforts
;Problem
: Neither the costs nor the benefits of using LD in Government is well known, let alone obvious.
;Goal
: Make the benefits of GLD obvious and memorable
;Solution Scope
:
;Participants
: [[User:Bandholtz|Thomas Bandholtz]] (3) - see [http://lists.w3.org/Archives/Public/public-egov-ig/2009Sep/0027.html]
;Status
: Just a vague idea that we should do something

;Community
: People trying to decide how much to support Government Linked Data (GLD) efforts
;Problem
: Neither the costs nor the benefits of using LD in Government is well known, let alone obvious.
;Goal
: Make the benefits of GLD obvious and memorable
;Solution Scope
: Performance Management and the Social Data Web
;Participants
: [[User:Gthomas2|George Thomas]] (it's part of what I'm doing at HHS :) see blurb on the [http://www.w3.org/2007/eGov/IG/wiki/Projects/SDW Social Data Web] (SDW) for the overall idea I'm going for. See [http://ontologies.socialdataweb.org example ontologies] and [http://govgoals.org example agency.gov/open instance data], and [http://george.thomas.name/hhs this presentation] for an overview and some screenshots that explain this wip.
;Status
: -
: I created [http://ontologies.socialdataweb.org/fedsheets//omb/ex53 OMB Exhibit53] and [http://ontologies.socialdataweb.org/fedsheets/omb/ex300 OMB Exhibit300] ontologies - these are very simple for demonstration purposes. They include one class and a bunch of datatype properties (that should be refactored) representing the way in which Feds use these spreadsheets and OMB collects this data, but with a first cut at some stronger typing and Web-centrism. I've also created Ex300 and Ex53 instance data for HHS using the [http://it.usaspending.gov US IT Dashboard], grabbing all the HHS Exhibit53 and Exhibit300 data from there, converting it to instances of the SDW OMB example ontologies.
: -
: I created a [http://www.omg.org/spec/BMM/1.0/ Business Motivation Model] ontology and published it on the 'Voluntary Consensus Standards Organization' [http://ontologies.socialdataweb.org/vcso/omg/bmm the SDW URI placeholder], which reflects the OMG concepts/terms and relationships standardized by that (pretty simple) open spec. I also created a bunch of BMM instance data, representing BMM instance models I've created that reflect some cloud computing goings on within the Federal CIO Council Cloud Computing Advisory Committee Workgroups that I'm participating in, and the Fed CIO Council Architecture and Infrastructure Committee where I co-chair the Services Subcommittee. I also created some example BMM data for HHS representing how they might leverage the CCAC and AIC projects. All of this instance data represents actual projects I'm working on, but for the purposes of this GLD demo, it's just a couple domain specific ontologies (capital planning and business motivation) along with associated instance data that get connected via another ontology, and everything just keeps building on the last thing...
: -
: I created a [http://ontologies.socialdataweb.org/bizmo Bizmo Ontology] as an example of an ontology that links these two domains, with the emphasis on the decentralized control that still allows for integration - the much abused 'federation' idea in gov circles is what I'm going for here. I created some Bizmo instances that leverages and integrates all the previously mentioned ontology specifications and their respective instance data from (the simulated) different organizational entities. Again, I'm not trying to demonstrate good ontology modeling, I'm just trying to demonstrate the characteristics of GLD on the Web, for comparing/contrasting with other approaches. It's a demo :)
: -
: I've imported all the ontologies and instance data published on the Web into a [http://http://semantic-mediawiki.org Semantic MediaWiki] using the [http://www.projecthalo.com/ Halo Extensions] as implemented by [http://wiki.ontoprise.de Ontoprise], currently running only on my Windows laptop localhost. This allows for ontology class/property driven faceted browsing, 'fact' based browsing, provides some query tools that are easy to use, a variety of export formats, RESTful RDF/XML GET'ing of anything while maintaining external vocab/type specs of internal articles, ontology class/property driven tag suggestion (or 'semantic annotations'), and more coming, like ontology driven forms (that I haven't fired up or played with yet). Since it's a Wiki, it features some of the SDW capabilities I'm after, with the idea being that Modelers provide the KB structures for knowledge workers to populate/curate etc. I'll soon get started working on standing up a [http://govgoals.org GovGOALS] (Government Goal-Outcome-Alignment-Line-of-Sight) instance of the SMW+Halo MediaWiki, either on my humble Ubuntu machine to make it accessible on the interwebs, or in some nicer environment - after working through the kinks on my Windows laptop (and hopefully getting through new ones introduced by this environment ;)
: -
: I plan to leverage other abstract and concrete modeling languages and formats, such as past and ongoing OMG work done in collaboration with OMB culminating in things like UML models of the Federal Enterprise Architecture, and will either use the [http://www.oegov.org/ oegov.org work from TopQuadrant], or create another SemWeb version that starts with the (extremely simple) OMG UML models. Some friends from modeldriven.org and topquadrant.com have graciously assisted in finessing the XMI exports from the [http://www.sparxsystems.com.au/ Sparx EA UML tool] (that the OMG GovDTF folks use) into the Eclipse EMF based (MOF'ish) [http://www.eclipse.org/modeling/mdt/?project=uml2tools UML tools] so that those of us lucky enough to have [http://topquadrant.com/products/TB_Composer.html really nice Onto tools] can do nice things with them. I've experimented with creating ontologies straight from the FEA Consolidated Reference Model XSD, but the 'structure preserving' transformation results were horrific, which IMHO is more of a reflection of the XSD's than the transformation algorithms. I will also probably look to leverage existing XSD's and instance data like Owen's StratML, most likely by transforming that data into BMM instance data, demonstrating XSD/XML integration (and design choice issues) with other standards like the BMM (insert standards joke here) so that the StratML data might be used as source data linked to other IT Dashboard data in a similar GLD fashion as I'm doing with Bizmo.
: -
: - I'll probably setup socialdataweb.org as a drupal site at some point, and experiment with its SemWeb and Social Media features too - would love to hear from any Drupal ninjas that might be interested in collaborating. I'm thinking that perhaps Drupal might be better at managing a community engaged in the creation/curation of metadata/data, while the SMW+ might be more advanced in the actual KB mgmt - but I don't know...
: -
: Initial Freebase demo info [http://www.w3.org/2007/eGov/IG/wiki/Projects/FB-PM-Demo here]
: I'd like to demo the same thing without using Freebase, with LOD style distributed ontologies and datasets using Open Source stuff
: Perhaps pick a few more concepts from different standards for integration
: Flesh out more real/dummy instance data

I've been using the contents of this [http://www.w3.org/2007/eGov/IG/wiki/Projects/SDW SDW] page (recreated from a US Fed Gov only wiki) as my best crack at simple language targeting Gov CIO types to describe the '''Social Data Web''' (with varying degrees of success, your discussion tab feedback appreciated). In 2009 I used the [http://www.freebase.com/labs/parallax/ Parallax] interface to [http://www.freebase.com/ Freebase] to demonstrate this, because it's so easy, to lots of Fed CoI's. [http://bizmo.freebase.com/ Bizmo] is a '''Performance Management''' (PM) starter example that I mocked up, initially to incorporate the [http://www.omg.org/spec/BMM/1.0/ OMG Business Motivation Model] and link that with some US Fed Gov 'Capital Planning and Investment Control' (CPIC) concepts. Basically it's just about showing how what you're spending on IT is helping you achieve your business goals (an oldie but goodie). This is a rich area, because there's a ton of prior art (architecture frameworks like FEA, FEAF, DODAF, MODAF, TOGAF, COSM/COBIT, RM-ODP, ZIFA, MDA, design and impl concerns like SOA/BPM/BRM, yadda yadda...) that would serve as interesting work to mine for simple expressions of perhaps just ten or twelve concepts from 'standards' such as these, and show how they relate to or augment each other.

Freebase is an easy way to demo the idea of socially designed and linking of light weight component ontologies, independently designed/linked by different communities of interest. Parallax is an easy way to showcase faceted search/browsing and the ability for users to traverse interlinked datasets without any apriori knowledge of data structure, and show off other interactive things like permalinking your filtered dataset and embedding that widget, etc. Some screenshots of how I've been demoing Parallax/Freebase are in a preso [http://www.slideshare.net/george.thomas.name/the-social-data-web here]. So while Freebase is a great example of, as the MetaWeb guys say a 'socially managed database', and that is a lot of the spirit I'mg going for here (especially how easy it is to manually create RDFS and the collaboration features around that) unfortunately we can't use it as an example of a SPARQL endpoint (only a MQL endpoint - which is cool, but obviously not optimal in the W3C context). If there are any [http://virtuoso.openlinksw.com/overview/VOSRelease/index.htm Virtuoso] gurus out there that want to participate (I've built/installed Virtuoso before, but that's about it), I'd love to put [http://sparallax.deri.ie/ DERI's Sparallax] in front of that to give us all the standards goodness with mostly the same Parallax features, and I know there are other social semantic tools out there as well for demonstrating social creation and linking of vocabularies (perhaps just Drupal or a Semantic MediaWiki would suffice? Not so hard to find social tools or semantic tools, but not a lot of these two together, at least not with level of Freebase that I know of) - please feel free to suggest others, with FLOSS of particular interest.

I think PM is a good topic because it's something all Gov Agencies have to do, and it is directly related to some top level US CIO goals (like the [http://it.usaspending.gov IT Dashboard]). Mechanistically a PM demo could also be used to exemplify other domain specific challenges, like the 'Comparative Effectiveness' objective that quantifies the 'Performance Based Care' goal in Healthcare. An example I use a lot in the Health Care domain (without any real deep/good thought) begins with the one you hear about - the ability to discover differences in health care costs in different locales - and while this is generally seen as an analysis of health care providers (which is a hot potato), it wouldn't be too much of a stretch (conceptually or technically) to imagine looking for causality in an EPA or NASA dataset (is there something in the water or the atmosphere that contributed to this health circumstance?). I hope this would further reinforce the idea of linking across the datasets being exposed by the [http://www.whitehouse.gov/blog/2009/12/08/promoting-transparency-government Open Government Directive] in the US and related showcase sites like data.gov in the US and (the much anticipated :-) data.gov.uk site and other awesome work from the UK Gov (and folks like JS!). I'd love to get GEO/GIS vocabs into the mix (as most Arch frameworks have a 'where' dimension) and ''need'' to get Licensing vocabs into the mix to demonstrate how IP is maintained. All for now! -g

== OGD Approval ==

;Community
: People inside government who want to implement Open Data systems
;Problem
: They can't get approval or resources for the projects: there is concern that they may have negative consequences for making data easily available to the public.
;Goal
: Have approval from the appropriate levels of (each) government, such that the folks with their hands on the data have management support for releasing it.
;Solution Scope
: Unclear. Is this covered by the above documents? Should we avoid lobbying, or assisting lobbyists?
;Participants
:
;Status
:Just a vague idea that we should do something

== GLD Approval ==

(Like "OGD Approval" but for Linked Data.)

;Community
: People inside government who want to implement Linked Data / Semantic Web systems
;Problem
: They can't get approval or resources for the projects: there is concern that they may have negative consequences for using certain W3C technologies
;Goal
: Approval/support from appropriate levels of governments, giving management support for the community
;Solution Scope
: See OGD approval. Consider activities like Tim Berners-Lee visiting the White House to talk about GLD (again).
;Participants
:
;Status
:Just a vague idea that we should do something

Projects

Dbennett — Wed, 10 Feb 2010 13:20:24 GMT

Dbennett:

'''DRAFTY, BUT GOOD ENOUGH. PLEASE SIGNUP USING [http://www.w3.org/2002/09/wbs/33280/project-signup/ Web Poll]'''

= About This Document =

== Project Template ==

For each project, there should be a table (as below) briefly answering certain key questions:

;Community
: (Who are the people directly affected by this project? Could also be called "target audience" or "users".)
;Problem
: (What problem is the community experiencing that we can help solve?)
;Goal
: (What will things be like when the problem is solved?)
;Solution Scope
: (What kinds of solutions will we be attempting or avoiding?)
;Participants
: (Who is willing to put in time to solving this problem? WikiNames are good here, and maybe put the number of hours per week you can commit in parens after your name)
;Status
: (What is currently going on with this project? What phase is it in?)

After this table, there is room for general discussion about the project.

== Terminology ==

'''OGD''' is Open Government Data, data that governments have that they are making available in machine-readable form (more than paper or PDFs) to the public. OGD typically uses XML, CSV, custom data formats, and/or Web APIs.

'''GLD''' is Government Linked Data, which is OGD that's using the Semantic Web standards (eg RDF), and also following the Linked Data architecture, where items are identified by URIs which lead to information about the items.

-----

= W3C eGov Projects (PROPOSED) =

== GLD Tech ==

;Community
:
;Problem
:
;Goal
:
;Solution Scope
: Create and publish an W3C Interest Group Note:
;Participants
: [[User:Bandholtz|Thomas Bandholtz]] (3) - see my posting [http://lists.w3.org/Archives/Public/public-egov-ig/2009Nov/0045.html] [[User:ChrisBeer|Chris Beer]] - happy to translate into "laymans terms" as well as add what I can
:
;Status
:Early conceptualization

'''Government Linked Data: Techniques and Technologies'''

Document description: Written for someone with a basic technical knowledge of deploying web servers, relational DBMS's, client-server architecture, etc. Assumes the reader is already convinced of the merits of LD.

Some agenda topics:
* how does linked data support (non-RDF) data consumers?
* strategies for modelling government data
* essential metadata for Government Linked Open Data (eg VoiD)
* expressing rights and licensing information
* approaches to provenance, authority and trust
* using RDF for Statistical Data

== GLD Strategy ==

;Community
:
;Problem
:
;Goal
:
;Solution Scope
: Create and publish an W3C Interest Group Note
;Participants
: [[User:Hughb|Hughb]] (1) - doing this one for the learning
: [[User:Bandholtz|Thomas Bandholtz]] (3) - some involvement in Germany and on the European level
: [http://ambur.net/bio.htm Owen Ambur] - My participation is conditional: If a set of measurable goals and objectives are outlined, I will render them in StratML format.
: [[User:SWebb2|Sylvia Webb]] (4) - Interested from ANSI ASC X12 perspective
;Status
:Early conceptualization

'''Government Linked Data: Strategies and Success Stories'''

Document description: A high-level planning/roadmap document, motivating the adoption of linked data technologies for government open data, and detailing the organizational steps to a successful deployment. Written for middle and high level IT managers. Does not assume the reader is convinced about using LD.

== Web Tech ==

;Community
: Staff of government entities looking implement or improve government services with Web technologies.
;Problem
:
;Goal
:
;Solution Scope
: Create and publish an W3C Interest Group Note
;Participants
: [[User:Hughb|Hughb]] (1) [[User:Briangryth|Briangryth]] (1) [[User:ChrisBeer|Chris Beer]] (5)
;Status
:Early conceptualization

'''Best Practices for Using Web Technologies to Deliver Government Services'''

Document description: Collection of everything we know about eGov, but not going into a lot of depth. Maybe a few technology details, but still usable for non-techies. The document will address both "Web 1.0" and "Web 2.0" technologies and how these technologies can help improve government services. The document will be useful to both IT staff and policy-makers and will direct the readers, when possible, to other materials for more in-depth information.

== Data Management ==

;Community
:
;Problem
:
;Goal
:
;Solution Scope
: Create and publish an W3C Interest Group Note
;Participants
: [[User:Hughb|Hughb]] (1)
: [[User:ChrisBeer|Chris Beer]] (2-3)
: [http://ambur.net/bio.htm Owen Ambur] - My participation is conditional, based upon the willingness of the group to address the need for a standard, minimum set of metadata elements in machine-readable (XML) format.
: [[User:SWebb2|Sylvia Webb]] (2-4)

;Status
:Early conceptualization

'''Best Practices for Government Data Management'''

Document description: Focussing on the particular issues of data, including interfacing with legacy systems and requirements for access. Some overlap with #1 and #2, but with a different emphasis for the folks whose greatest interest is in long-term stability of data.

== Social Media ==

;Community
: Government employees, at all levels, who would like to use social media tools and services.
;Problem
: Justifying the use of such services in light of the security risks and records management problems associated with them
;Goal
: To collaborate with other organizations and groups to provide and outline the best practices for using such services to advance the mission of government agencies and engage the public. The best will also highlight how security risks can be mitigated and records management requirements can be met.
;Solution Scope
: Create and publish an W3C Interest Group Note
;Participants
:[[User:Briangryth|Briangryth]] (1-3) [[User:ChrisBeer|Chris Beer]] (5-10) [http://ambur.net/bio.htm Owen Ambur]
;Status
:Early conceptualization

'''Issues and Best Practices in Government Use of Social Media'''

Document description: Propriety, ethics, how officials ought to engage, .... everyone else is already talking about it already, but let's try to bring what we can together with our international, standards-based perspective. Usage of such services should be directly linked to (aligned with) objectives documented in agency strategic plans and employee performance plans. The document should also address:

* How social media advances the mission of government?
* How social media usage can increases efficiency and productivity?
* How social media tools can support the implementation OGD and GLD?
* How social media tools can interfaces and support current government Web presences?

== GLD Demo ==

;Community
: People trying to decide how much to support Government Linked Data (GLD) efforts
;Problem
: Neither the costs nor the benefits of using LD in Government is well known, let alone obvious.
;Goal
: Make the benefits of GLD obvious and memorable
;Solution Scope
:
;Participants
: [[User:Bandholtz|Thomas Bandholtz]] (3) - see [http://lists.w3.org/Archives/Public/public-egov-ig/2009Sep/0027.html]
;Status
: Just a vague idea that we should do something

;Community
: People trying to decide how much to support Government Linked Data (GLD) efforts
;Problem
: Neither the costs nor the benefits of using LD in Government is well known, let alone obvious.
;Goal
: Make the benefits of GLD obvious and memorable
;Solution Scope
: Performance Management and the Social Data Web
;Participants
: [[User:Gthomas2|George Thomas]] (it's part of what I'm doing at HHS :) see blurb on the [http://www.w3.org/2007/eGov/IG/wiki/Projects/SDW Social Data Web] (SDW) for the overall idea I'm going for. See [http://ontologies.socialdataweb.org example ontologies] and [http://govgoals.org example agency.gov/open instance data], and [http://george.thomas.name/hhs this presentation] for an overview and some screenshots that explain this wip.
;Status
: -
: I created [http://ontologies.socialdataweb.org/fedsheets//omb/ex53 OMB Exhibit53] and [http://ontologies.socialdataweb.org/fedsheets/omb/ex300 OMB Exhibit300] ontologies - these are very simple for demonstration purposes. They include one class and a bunch of datatype properties (that should be refactored) representing the way in which Feds use these spreadsheets and OMB collects this data, but with a first cut at some stronger typing and Web-centrism. I've also created Ex300 and Ex53 instance data for HHS using the [http://it.usaspending.gov US IT Dashboard], grabbing all the HHS Exhibit53 and Exhibit300 data from there, converting it to instances of the SDW OMB example ontologies.
: -
: I created a [http://www.omg.org/spec/BMM/1.0/ Business Motivation Model] ontology and published it on the 'Voluntary Consensus Standards Organization' [http://ontologies.socialdataweb.org/vcso/omg/bmm the SDW URI placeholder], which reflects the OMG concepts/terms and relationships standardized by that (pretty simple) open spec. I also created a bunch of BMM instance data, representing BMM instance models I've created that reflect some cloud computing goings on within the Federal CIO Council Cloud Computing Advisory Committee Workgroups that I'm participating in, and the Fed CIO Council Architecture and Infrastructure Committee where I co-chair the Services Subcommittee. I also created some example BMM data for HHS representing how they might leverage the CCAC and AIC projects. All of this instance data represents actual projects I'm working on, but for the purposes of this GLD demo, it's just a couple domain specific ontologies (capital planning and business motivation) along with associated instance data that get connected via another ontology, and everything just keeps building on the last thing...
: -
: I created a [http://ontologies.socialdataweb.org/bizmo Bizmo Ontology] as an example of an ontology that links these two domains, with the emphasis on the decentralized control that still allows for integration - the much abused 'federation' idea in gov circles is what I'm going for here. I created some Bizmo instances that leverages and integrates all the previously mentioned ontology specifications and their respective instance data from (the simulated) different organizational entities. Again, I'm not trying to demonstrate good ontology modeling, I'm just trying to demonstrate the characteristics of GLD on the Web, for comparing/contrasting with other approaches. It's a demo :)
: -
: I've imported all the ontologies and instance data published on the Web into a [http://http://semantic-mediawiki.org Semantic MediaWiki] using the [http://www.projecthalo.com/ Halo Extensions] as implemented by [http://wiki.ontoprise.de Ontoprise], currently running only on my Windows laptop localhost. This allows for ontology class/property driven faceted browsing, 'fact' based browsing, provides some query tools that are easy to use, a variety of export formats, RESTful RDF/XML GET'ing of anything while maintaining external vocab/type specs of internal articles, ontology class/property driven tag suggestion (or 'semantic annotations'), and more coming, like ontology driven forms (that I haven't fired up or played with yet). Since it's a Wiki, it features some of the SDW capabilities I'm after, with the idea being that Modelers provide the KB structures for knowledge workers to populate/curate etc. I'll soon get started working on standing up a [http://govgoals.org GovGOALS] (Government Goal-Outcome-Alignment-Line-of-Sight) instance of the SMW+Halo MediaWiki, either on my humble Ubuntu machine to make it accessible on the interwebs, or in some nicer environment - after working through the kinks on my Windows laptop (and hopefully getting through new ones introduced by this environment ;)
: -
: I plan to leverage other abstract and concrete modeling languages and formats, such as past and ongoing OMG work done in collaboration with OMB culminating in things like UML models of the Federal Enterprise Architecture, and will either use the [http://www.oegov.org/ oegov.org work from TopQuadrant], or create another SemWeb version that starts with the (extremely simple) OMG UML models. Some friends from modeldriven.org and topquadrant.com have graciously assisted in finessing the XMI exports from the [http://www.sparxsystems.com.au/ Sparx EA UML tool] (that the OMG GovDTF folks use) into the Eclipse EMF based (MOF'ish) [http://www.eclipse.org/modeling/mdt/?project=uml2tools UML tools] so that those of us lucky enough to have [http://topquadrant.com/products/TB_Composer.html really nice Onto tools] can do nice things with them. I've experimented with creating ontologies straight from the FEA Consolidated Reference Model XSD, but the 'structure preserving' transformation results were horrific, which IMHO is more of a reflection of the XSD's than the transformation algorithms. I will also probably look to leverage existing XSD's and instance data like Owen's StratML, most likely by transforming that data into BMM instance data, demonstrating XSD/XML integration (and design choice issues) with other standards like the BMM (insert standards joke here) so that the StratML data might be used as source data linked to other IT Dashboard data in a similar GLD fashion as I'm doing with Bizmo.
: -
: - I'll probably setup socialdataweb.org as a drupal site at some point, and experiment with its SemWeb and Social Media features too - would love to hear from any Drupal ninjas that might be interested in collaborating. I'm thinking that perhaps Drupal might be better at managing a community engaged in the creation/curation of metadata/data, while the SMW+ might be more advanced in the actual KB mgmt - but I don't know...
: -
: Initial Freebase demo info [http://www.w3.org/2007/eGov/IG/wiki/Projects/FB-PM-Demo here]
: I'd like to demo the same thing without using Freebase, with LOD style distributed ontologies and datasets using Open Source stuff
: Perhaps pick a few more concepts from different standards for integration
: Flesh out more real/dummy instance data

I've been using the contents of this [http://www.w3.org/2007/eGov/IG/wiki/Projects/SDW SDW] page (recreated from a US Fed Gov only wiki) as my best crack at simple language targeting Gov CIO types to describe the '''Social Data Web''' (with varying degrees of success, your discussion tab feedback appreciated). In 2009 I used the [http://www.freebase.com/labs/parallax/ Parallax] interface to [http://www.freebase.com/ Freebase] to demonstrate this, because it's so easy, to lots of Fed CoI's. [http://bizmo.freebase.com/ Bizmo] is a '''Performance Management''' (PM) starter example that I mocked up, initially to incorporate the [http://www.omg.org/spec/BMM/1.0/ OMG Business Motivation Model] and link that with some US Fed Gov 'Capital Planning and Investment Control' (CPIC) concepts. Basically it's just about showing how what you're spending on IT is helping you achieve your business goals (an oldie but goodie). This is a rich area, because there's a ton of prior art (architecture frameworks like FEA, FEAF, DODAF, MODAF, TOGAF, COSM/COBIT, RM-ODP, ZIFA, MDA, design and impl concerns like SOA/BPM/BRM, yadda yadda...) that would serve as interesting work to mine for simple expressions of perhaps just ten or twelve concepts from 'standards' such as these, and show how they relate to or augment each other.

Freebase is an easy way to demo the idea of socially designed and linking of light weight component ontologies, independently designed/linked by different communities of interest. Parallax is an easy way to showcase faceted search/browsing and the ability for users to traverse interlinked datasets without any apriori knowledge of data structure, and show off other interactive things like permalinking your filtered dataset and embedding that widget, etc. Some screenshots of how I've been demoing Parallax/Freebase are in a preso [http://www.slideshare.net/george.thomas.name/the-social-data-web here]. So while Freebase is a great example of, as the MetaWeb guys say a 'socially managed database', and that is a lot of the spirit I'mg going for here (especially how easy it is to manually create RDFS and the collaboration features around that) unfortunately we can't use it as an example of a SPARQL endpoint (only a MQL endpoint - which is cool, but obviously not optimal in the W3C context). If there are any [http://virtuoso.openlinksw.com/overview/VOSRelease/index.htm Virtuoso] gurus out there that want to participate (I've built/installed Virtuoso before, but that's about it), I'd love to put [http://sparallax.deri.ie/ DERI's Sparallax] in front of that to give us all the standards goodness with mostly the same Parallax features, and I know there are other social semantic tools out there as well for demonstrating social creation and linking of vocabularies (perhaps just Drupal or a Semantic MediaWiki would suffice? Not so hard to find social tools or semantic tools, but not a lot of these two together, at least not with level of Freebase that I know of) - please feel free to suggest others, with FLOSS of particular interest.

I think PM is a good topic because it's something all Gov Agencies have to do, and it is directly related to some top level US CIO goals (like the [http://it.usaspending.gov IT Dashboard]). Mechanistically a PM demo could also be used to exemplify other domain specific challenges, like the 'Comparative Effectiveness' objective that quantifies the 'Performance Based Care' goal in Healthcare. An example I use a lot in the Health Care domain (without any real deep/good thought) begins with the one you hear about - the ability to discover differences in health care costs in different locales - and while this is generally seen as an analysis of health care providers (which is a hot potato), it wouldn't be too much of a stretch (conceptually or technically) to imagine looking for causality in an EPA or NASA dataset (is there something in the water or the atmosphere that contributed to this health circumstance?). I hope this would further reinforce the idea of linking across the datasets being exposed by the [http://www.whitehouse.gov/blog/2009/12/08/promoting-transparency-government Open Government Directive] in the US and related showcase sites like data.gov in the US and (the much anticipated :-) data.gov.uk site and other awesome work from the UK Gov (and folks like JS!). I'd love to get GEO/GIS vocabs into the mix (as most Arch frameworks have a 'where' dimension) and ''need'' to get Licensing vocabs into the mix to demonstrate how IP is maintained. All for now! -g

== OGD Approval ==

;Community
: People inside government who want to implement Open Data systems
;Problem
: They can't get approval or resources for the projects: there is concern that they may have negative consequences for making data easily available to the public.
;Goal
: Have approval from the appropriate levels of (each) government, such that the folks with their hands on the data have management support for releasing it.
;Solution Scope
: Unclear. Is this covered by the above documents? Should we avoid lobbying, or assisting lobbyists?
;Participants
:
;Status
:Just a vague idea that we should do something

== GLD Approval ==

(Like "OGD Approval" but for Linked Data.)

;Community
: People inside government who want to implement Linked Data / Semantic Web systems
;Problem
: They can't get approval or resources for the projects: there is concern that they may have negative consequences for using certain W3C technologies
;Goal
: Approval/support from appropriate levels of governments, giving management support for the community
;Solution Scope
: See OGD approval. Consider activities like Tim Berners-Lee visiting the White House to talk about GLD (again).
;Participants
:
;Status
:Just a vague idea that we should do something