Re: [ontolog-forum] Making the Ontology Summit content Accessible

[Ali Hashemi <ali@xxxxxxxxx>]

Hi Peter,

On Fri, Aug 12, 2011 at 1:40 PM, Peter Yim <peter.yim@xxxxxxxx> wrote:

[ppy]  (ref. http://ontolog.cim3.net/forum/ontolog-forum/2011-07/msg00184.html#nid05
) I assume you meant "purple semantic media wiki (PSMW)," as we are
actually working on moving out of "purple wiki." (You probably meant
that already, I just wanted to make sure.)

Yes, I meant PSMW.

 

> [AH]   Do you think a wiki is the best fit for a final website ...

[ppy]  <snip> (ii) more modern
wiki platforms ("psmw" can be considered among them) are so versatile
that  one can support a lot of fancy graphics and GUI's (to the extent
that most people may not even realize that some websites are built on
wiki platforms) ... as such, the decision should not be made on
whether or not the platform is labeled as a "wiki," but rather, on
what functionalities and features are needed in your case, and what
platform would most easily provide for them.

Absolutely agree. My main objection to a wiki was exactly the lack of that versatility that you mention is now available. Going to a site that has the look and feel of a wiki is good for some things, but wholly inappropriate for many others. Bart and I in our initial pass, identified 4 broad classes of users.

Again, keeping in mind (and assuming) that the function of the website is to support the content and end result of each summit, the functionality that arose for each class of user were:

1. General browser - "What was the summit about?"
2. Evangelists - Someone wanting to make the case - "How can I make the case to my specific audience?"
3. Decision Makers - Someone looking for ontology uses - "Which efforts match my current situation?"
4. Contributors - "I have an experience that fits this summit"

I think the Contributor class corresponds to the classic wiki use-case, the other 3 classes fall more into the consumer use-case.

[ppy]  this is great! ... except for the word "expressive," I bet PSMW
would be a good candidate to support the work. As for "expressive"
(especially "expressive ontology"), PSMW may possibly be challenged;
we need to talk things through together, and probably work something
out together. I am sure to get to where you want (regardless of the
platform you end up choosing), some code have yet to be developed, and
that's where some of the gaps may be bridged.

 

Good point, but there's a reason for choosing an expressive ontology. We do not foresee the more expressive version as one that would necessarily be computed upon. We are actually hoping to use this project as a way to demonstrate two types of ontology use - that of ontology for design, and ontology for real-time search / usability. To that end, we're thinking of deploying two distinct ontology artifacts. One proposal we're considering for how to move forward and the implementation architecture.

1. Develop an expressive, though under-specified ontology. This captures in a technology independent way, what we're trying to solve / represent. (candidate language - Common Logic)
2. Tag our content using the vocabulary we've developed (candidate language, RDFa or some other microformat)
3. Determine the types of inferences and views we want to support.
4. Create and store queries that correspond to these inferences using our RDFa tagging. (candidate query languages, SPARQL or if we want to go the old school route, perl scripts?)
5. Develop a set of CSS's that provide the visual context and style information for how our content is displayed.

So there are a number of interesting uses of ontology that come out of this. The ideas that have driven the above approach include:

• It is easier to conduct semantic mappings on expressive ontologies
• Demonstrate how ontology driven design can work with a practical (if simple) example.
• Demonstrate how an expressive, referent ontology can provide as a sort of "contract store", whereby various web services can align or rather honour the semantics of a referent ontology in their functionality -- this is entirely in alignment with Service Oriented Architectures which are coming to dominate web application deployments. 
• I personally think that Ontology Driven Software / Web Development is an area that is all too often overlooked (notwithstanding the excellent conference on that very topic).
• The link between SOA and ontology is also not well highlighted, which this project could help clarify
• Demonstrate an effective use of  query-rewriting to match the technology choice of du jour.
• Demonstrate how the Linked Open Data world can be enriched and enhanced by linking the deployed microformats or triples to an actual, expressive ontology.

To this end, we actually took a close look at the various available extensions to SemanticMediaWiki, and our main issue was the lack of versatility when it comes to rendering the pages. 

 

> [AH]  Would we be able to store say a triple-store on ontolog?

[ppy]  I would recommend using OOR (ref.
http://ontolog.cim3.net/cgi-bin/wiki.pl?OpenOntologyRepository )

If the PSWM or Ontolog is already deploying such a thing, then we don't need a distinct space. 

The idea however, is to annotate the content of the Summit - wherever it is stored - with a microformat vocabulary derived from the more expressive referent ontology. Queries will be rewritten according to common inferences based on the referent ontology, and deployed in the website. The main functionality we desire, is to render the results according to our own CSS's and not be tied to the CSS's that come with a particular SMW implementation. If this freedom is available, and if PSMW is being deployed anyway, then it does not make sense to have a distinct annotated store.

Of course, aside from Bart and yourself (and some preliminary discussions at the Summit debriefing meeting a few months back), precious little feedback has trickled in. So all this could be subject to change if anyone else has a better idea or vision for the architecture or execution.

Best,

Ali 

On Fri, Aug 12, 2011 at 10:13 AM, Ali Hashemi <ali@xxxxxxxxx> wrote:
> Dear Peter and Bart,
> Sorry for the delayed response.I would definitely be interested in the work
> on the purple wiki's, though I don't have much time to commit to that.
> In the meantime, Bart G. and I have started work on this idea. Do you think
> a wiki is the best fit for a final website associated with each year's
> summit? Are people on board with adding an additional, explicit deliverable
> from each summit (i.e. a website that summarizes and provides an easy, more
> general audience view into the work)?
> For this current project, we were interested in using an expressive ontology
> to capture a technology agnostic representation of the domain, and then do a
> sort of "ontology-driven website design", deploying suitable, more optimized
> semantic technologies for the actual website functionality.
> Would we be able to store say a triple-store on ontolog? We're a bit unclear
> how the web infrastructure would work on this.
> Best,
> Ali


> On Wed, Jul 27, 2011 at 7:56 PM, Peter Yim <peter.yim@xxxxxxxx> wrote:
>>
>> Sounds great, Ali!
>>
>> > [AH]  (micro-tears on ankle ligaments means no camping+cottaging for a
>> > few weeks! :D ).
>>
>> [ppy]  get well soon, Ali!
>>
>>
>> In addition to your thoughts, I want to bring up the following for you
>> and everyone's consideration ...
>>
>> 1.  There is an effort (behind the scene) that is working on
>> "migrating and enhancing" the OntologWiki content -- transforming that
>> from its current "purple wiki" platform to a "purple semantic
>> mediawiki (PSMW)" platform.
>>
>> 2.  presumably one could be a bit more fancy with the "look and feel"
>> on that new platform as well. Do check out:
>>
>> http://ontologydesignpatterns.org/wiki/Ontology_Design_Patterns_._org_%28ODP%29
>> and
>> http://smwforum.ontoprise.com/smwforum/index.php/HaloExtension_Product_Information
>> for possibilities (on how the site may be implemented) ... both of
>> these sites are build on the Semantic MediaWiki platform.
>>
>> 3.  you (and others who are interested to support this initiative you
>> are brining up) might consider possible synergies, through
>> collaborating with those who are working on the PSMW "migration and
>> enhancement" now (these people are meeting Wednesday evenings every
>> week - I can put you in touch with them.)
>>
>>
>> Thanks & regards.  =ppy
>> --


>> ---------- original message ----------
>> From: Ali Hashemi <ali@xxxxxxxxx>
>> Date: Wed, Jul 27, 2011 at 7:22 AM
>> Subject: Re: Making the Ontology Summit content Accessible
>> To: Peter Yim <peter.yim@xxxxxxxx>, "[ontolog-forum]"
>> <ontolog-forum@xxxxxxxxxxxxxxxx>, Michael F Uschold
>> <uschold@xxxxxxxxx>
>>
>>
>> Hi Peter and all,
>>
>> After a long delay, I am finally capable of committing some time and
>> energy on this effort (micro-tears on ankle ligaments means no
>> camping+cottaging for a few weeks! :D ).
>>
>>
>> To briefly recap what's written below.
>>
>> Ontology Summit 2011 came and went, and was by many measures a
>> success. Many wonderful conversations were stimulated and captured on
>> ontolog wiki and the Communique with its supporting documentation
>> provides a valuable resource to those who wish to make the case for
>> ontology. However, it was noticed by many that the content is not as
>> accessible to a broad audience as we'd like. Moreover, I think it was
>> also noticed that we could derive additional value from the summit by
>> creating a distinct, dedicated website for the work that was developed
>> in the summit.
>>
>> As Michael Uschold and others noted, the way that content is presented
>> on the ontolog wiki is well geared for collaboration, editing and
>> creating, but not great for a broader audience. To address this issue,
>> I think a proposal is to have the creation of a Summit Website to be
>> an explicit goal of each summit from here on in.
>>
>> To put our efforts where our mouths are, I think we can begin by
>> creating a site for the Making the Case. Specifically, I think we can
>> achieve three objectives in so doing:
>>
>> 1. Present our work in engaging, custom-way way to a broad audience
>>
>> 2. Actually use ontology to do so
>>
>>  2.1  Demonstrate the use of an expressive, technology agnostic,
>> though lightweight ontology in the conceptualization of our problem
>> domain
>>
>>  2.2  Demonstrate the mapping of the expressive ontology to a
>> specific technology implementation that fits the needs of the problem
>> domain
>>
>> Where (2.1) and (2.2) provide two additional case examples of ontology
>> in action.
>>
>>
>> To wit, in making the case, we considered:
>>
>> o  Target audiences
>> o  Values and benefits
>> o  Case studies
>> o  Metrics
>>
>> these each comprised separate tracks, yet any person who wants to use
>> the output of the summit will likely be looking for a specific
>> audience, where that audience values a subset of the metrics and a
>> subset of the benefits, where in addition, only a subset of the case
>> studies actually demonstrate these benefits. So by modeling how the
>> output of the summit fits into the work ecology of our potential
>> audience, we can structure our content in such a way to generate
>> dynamic, custom views to the website users. Moreover, if this summit
>> continues to collect more case studies, we would be able to
>> organically grow the site as we add content with minimal costs.
>> All-in-all, a good demonstration of what we're preaching about
>> ontology, applied to itself - no?
>>
>> I think (2.1) and (2.2) above demonstrate two uses of ontologies. The
>> first (2.1) captures ontology in a technology agnostic way in an
>> expressive language. I would propose using Common Logic to capture the
>> relationship among the various tracks and our users. The second
>> ontology use (2.2) corresponds to using the tools that are available
>> now -- this involves mapping elements of minimally-defined CL theory
>> to RDF/XML(?) statements that annotate our content + (stored) SPARQL
>> queries that are used to generate the views according to user input.
>> The idea is that the user won't be explicitly writing SPARQL queries,
>> but rather, a nice, clean interface that elicits their specific needs
>> as it pertains to making the case, would execute the queries and
>> generate a dynamic, custom-tailored page.
>>
>>
>> What do people think? Any feedback?
>>
>> And more importantly, are there people who would be willing to donate
>> some time to realize this effort? I know Michael U. had wanted to
>> refine our metrics and specifically the benefits that each use case
>> purported to demonstrate. What we need for this effort are:
>>
>> o  Refined metrics for the metric tracks
>> o  Standardized benefits in terms of the metrics for each of the use cases
>> o  Linking the target audiences to types of benefits
>> o  Minimally specified CL modules that represent the relations between
>> the various tracks
>> o  A mapping of fragments of the CL modules to RDF annotations and
>> SPARQL queries
>> o  An actual website that implements SPARQL and RDF/XML
>>
>> Any volunteers for any of these tasks? Also, please note that CL, RDF
>> and SPARQL are just initial proposals to get the ball rolling, if
>> others feel that something else is more appropriate, please speak up.
>>
>> Thanks in advance,
>> Ali


>> On Wed, May 25, 2011 at 10:09 AM, Peter Yim <peter.yim@xxxxxxxx> wrote:
>> >
>> > Great, Ali! ... Forwarding this thread to the Ontolog community now ...
>> >
>> > ALL: Let's continue the conversation here (on the [ontolog-forum] list).
>> >
>> > Thanks & regards.  =ppy
>> > --


>> > ---------- Forwarded message ----------
>> > From: Ali SH <asaegyn+out@xxxxxxxxx>
>> > Date: Wed, May 25, 2011 at 5:26 AM
>> > Subject: Re: [ontology-summit] Making the Summit Accessible
>> > To: Ontology Summit 2011 discussion <ontology-summit@xxxxxxxxxxxxxxxx>
>> >
>> >
>> > Dear all,
>> > Yes, it seems like a good idea to shift the conversation over to the
>> > regular forum.
>> > Unfortunately, I will be without electronic communication until June
>> > 6, so I won't be able to contribute until then.
>> > I will catch up then.
>> > Best,
>> > Ali
>> >
>> >
>> > --
>> >
>> >
>> > (•`'·.¸(`'·.¸(•)¸.·'´)¸.·'´•) .,.,


>> > On Sun, May 22, 2011 at 2:35 PM, Peter Yim <peter.yim@xxxxxxxx> wrote:
>> > >
>> > > Dear Ali and All,
>> > >
>> > >
>> > > Thank you, Ali ... this is great! ... I concur and encourage everyone
>> > > interested to actively engage in this conversation.
>> > >
>> > > 1.  I agree with you, and am in favor of having *one website* as the
>> > > home to all summit "presentations" year-after-year. (I believe that
>> > > academic conferences have their annual conferences distributed in
>> > > different site-locations are a result of how "ownerships" get passed
>> > > from institution to institution, and not a function of optimal
>> > > design.)
>> > >
>> > > 2.  note that the Ontolog-CWE (collaborative work environment)
>> > > actually have four key components in the infrastructure (a
>> > > portal/website space, a wiki, an archived mailing list and a webdav
>> > > server ... representing four somewhat orthogonal workspaces - a
>> > > presentation space, a collaborative authoring and synchronization
>> > > workspace, a conversation space, and a shared-file repository.) With
>> > > your effort here, looks like we can finally take advantage of the
>> > > portal/website infrastructure that has been sitting around all these
>> > > years.
>> > >
>> > > 3.  since OntologySummit2011 is officially over, and this exercise
>> > > that you are leading is actually using OntologySummit2011 as a case to
>> > > develop something that extends beyond this year's Summit and is
>> > > important to the entire ontology community, you might consider moving
>> > > the conversation to the [ontolog-forum] list, where the reach is wider
>> > > (roughly twice the number of subscribers, and more international
>> > > participation.)
>> > >
>> > > 4.  to augment this threaded discussion, please consider picking one
>> > > (or even several) time slots to run real-time focused discussion
>> > > and/or workshop(s) on this effort, making use of, say, the regular
>> > > Ontolog Thursday event time slot and virtual panel discussion session
>> > > format, as you feel appropriate. Reserve any date that is marked
>> > > "open" on our Ontolog master event calendar - see:
>> > > http://ontolog.cim3.net/cgi-bin/wiki.pl?MeetingsCalls (email me if I
>> > > can be of help to facilitate the organization of such event(s).)
>> > >
>> > >
>> > > Thanks & regards.  =ppy
>> > > --


>> > > On Thu, May 19, 2011 at 9:53 PM, Ali Hashemi <ali@xxxxxxxxx> wrote:
>> > > > beyond the conclusion of the face-to-face meetings.
>> > > > As a follow up to yesterday's conference call
>> > > > (http://ontolog.cim3.net/cgi-bin/wiki.pl?ConferenceCall_2011_05_19),
>> > > > I think
>> > > > we agreed on the need for developing something more than a
>> > > > Communique. We
>> > > > need to present the culmination of the summit (Communique + Tracks +
>> > > > Wiki
>> > > > content) in a more effective manner.
>> > > >
>> > > > Specifically, if we're considering putting the creation of a website
>> > > > as an
>> > > > additional explicit goal of the outcome of future summits, then I
>> > > > think we
>> > > > have one of two choices:
>> > > >
>> > > > One central site that contains each year
>> > > > One site for each year (i.e. how academic conferences usually
>> > > > collect
>> > > > material)
>> > > >
>> > > > I think the first one makes more sense, as it provides a more
>> > > > unified view
>> > > > of the progress of ontology and the summits. My personal experience
>> > > > with
>> > > > conference websites (say for IJCAI) is that each year differs highly
>> > > > in
>> > > > quality, they are not presented in a consistent way, and are
>> > > > generally a
>> > > > frustrating way to keep track of conferences over a long period of
>> > > > time.
>> > > > Beyond the above consideration, I would suggest that the purpose of
>> > > > each
>> > > > site should be to support the theme of the summit and mediate the
>> > > > relation
>> > > > to resources developed over the course of the summit in a more
>> > > > accessible
>> > > > manner.
>> > > > I'll use the 2011 Making the Case Summit to illustrate what I mean
>> > > > by the
>> > > > above statement.
>> > > > In this case, we identified a number of tracks tackling different
>> > > > aspects of
>> > > > one problem -- how to construct a compelling, persuasive argument re
>> > > > ontologies. In the course of this process, we collected, developed
>> > > > and are
>> > > > ultimately providing the material for ontology evangelists to make
>> > > > actual
>> > > > cases. Not only that, but the resources we provide include
>> > > > identifying a set
>> > > > of target audiences and broad strategies that evangelists might
>> > > > actually
>> > > > employ.
>> > > > !!
>> > > > The fact that an ontology evangelist would use the output of the
>> > > > summit to
>> > > > make a case should drive our organization and access to the
>> > > > collected and
>> > > > developed material. That is how a site would support the theme of
>> > > > this
>> > > > year's summit.
>> > > > To briefly recap,
>> > > >
>> > > > We identified a number of different audiences
>> > > >
>> > > > who care about a number of different metrics
>> > > >
>> > > > We identified a set of benefits that ontology can provide
>> > > >
>> > > > with corersponding metrics
>> > > >
>> > > > We solicited and collected a number of use cases
>> > > >
>> > > > where presumably, ontology actually delivered those benefits
>> > > > and it is expressible via the metrics.
>> > > >
>> > > > Remembering why an evangelist would be accessing the communique in
>> > > > the first
>> > > > place, this suggests a natural layout... Just to be explicit, an
>> > > > ontology
>> > > > evangelist wants to persuade at least the audiences we identified
>> > > > (+perhaps
>> > > > others that we missed) using at least the resources we provided. So
>> > > > given
>> > > > their audience, they’re interested in only a subset of the benefits,
>> > > > metrics
>> > > > and use cases at any one time. Moreover it would be useful for them
>> > > > to see
>> > > > which use cases and value metrics apply to which audience member.
>> > > > So... We should capture these relations in our content, and provide
>> > > > views
>> > > > into the summit web site according to the evangelist's target
>> > > > audience.
>> > > >
>> > > > (Evangelist (wants_to) convince TargetAudience)
>> > > > (TargetAudiences value Benefits)
>> > > > (TargetAudiences respond_to Metrics)
>> > > > (Metrics measure Benefits)
>> > > > (UseCases deliver Benefits)
>> > > >
>> > > > The ValueMetrics Synthesis (
>> > > >
>> > > > http://ontolog.cim3.net/cgi-bin/wiki.pl?OntologySummit2011_ValueMetrics_Synthesis
>> > > > ) already largely captures the mapping between the UseCases and both
>> > > > Benefits and Metrics.
>> > > > As Michael Uschold noted in today’s meeting, we should be able to
>> > > > develop an
>> > > > ontology for the usage framework. I believe it is also possible to
>> > > > connect
>> > > > that with the value metrics, and finally connect that to the target
>> > > > audience
>> > > > to create a tight loop to drive the development of our web effort.
>> > > > What we need to do is make these relations a bit more formal (and
>> > > > perhaps
>> > > > machine readable)!  And also, clearly articulate which Benefits and
>> > > > which
>> > > > ValueMetrics correspond to which TargetAudience. Machine readable
>> > > > representations are particularly desirably if we want to grow the
>> > > > usage
>> > > > example collection and provide dynamic views of our resources to the
>> > > > users.
>> > > > With such a structure in place, we can then develop a site that
>> > > > better
>> > > > corresponds to evangelist needs. Though of course, it would also be
>> > > > useful
>> > > > to have a presentation scheme that presents the story of the
>> > > > evolution of
>> > > > the summit as well.
>> > > > Are there any volunteers? Might someone in the ValueSynthesis track
>> > > > be able
>> > > > to extract the relevant bits of the matrix in some formalism? Can we
>> > > > agree
>> > > > on a vocabulary for audience, benefits, metrics and use case types
>> > > > in a
>> > > > machine readable way? <-- This is already informally done in the
>> > > > communique+tracks to some degree. The results of this analysis will
>> > > > at the
>> > > > very least drive the layout of the pages+views, and perhaps
>> > > > facilitate the
>> > > > technology implementation for the delivery of "nuggets" of content
>> > > > that
>> > > > we'll be hosting. We can discuss what a "nugget" of content means
>> > > > for this
>> > > > summit...
>> > > > Best,
>> > > > Ali
>> > > >
>> > > >
>> > > > --
>> > > >
>> > > >
>> > > > (•`'·.¸(`'·.¸(•)¸.·'´)¸.·'´•) .,.,

--


(•`'·.¸(`'·.¸(•)¸.·'´)¸.·'´•) .,.,

_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/  
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/  
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/ 
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J    (01)