Re: What is the SIOC-o-sphere?

Ivan Herman wrote:
> Wing C Yung wrote:
>   
>> I think that it's great that we have some SIOC producers; it's definitely
>> the first step towards *enabling* a potential "killer app" for the Semantic
>> Web. Having the data as RDF is necessary but not sufficient to convince
>> people of the benefits. The "killer app" is going to be a consumer of data
>> (like a SW-enabled feed aggregator/reader that does something compelling
>> with the data), not the data itself. Right now, are there other consumers
>> besides the one [1] listed on the SIOC page?
>>
>> [1]
>> http://sparql.captsolo.net/browser/browser.py?url=http%3A%2F%2Fwww.johnbreslin.com%2Fblog%2Findex.php%3Fsioc_type%3Dpost%26sioc_id%3D462
>>
>> Wing
>>
>>     
>
> It depends on the community, of course, but generally I concur.
>
> Let me tell you about ne specific experience I had (Karen was there,
> too, she probably remembers). We, ie, some team people from W3C, had a
> meeting with a whole group of *technical* people (not manager types)
> from Merck. It was a very long meeting; it included a presentation on
> Piggy Bank, some discussion with Tim, and a very long
> presentation/discussion led by me. Actually, it was quite successful;
> Merck joined W3C and are now active participants of the HCLS IG.
>
> As part of the presentation I also used the tabulator to show the power
> of merging/integrating data. Without entering into details, it used all
> kinds of data spread around on W3C site and elswhere on various
> technical documents, their translations, the persons involved and their
> data, etc. Sure, the tabulator does not have the smoothest interface,
> but the point could get across.
>
> After the presentation, the leader of the Merck delegation (a very
> friendly chap whom I knew because he had been at my tutorial in
> Edinburgh earlier that year) came to me and said privately: that type of
> demonstration does not work for them. It is way too geeky, it is way too
> much the 'same story' that they hear, but it does not really show what
> this technology can do for *them*. What made them move is the
> realization of using their data in their databases, integrate those,
> combined with their own real problems that they encounter.
>
> To avoid misunderstandings: SIOC is a great stuff (I have the radar
> running on my Firefox:-). That is not the point. The point is that for
> *corporate developers* we need specific use cases in their own
> environments and settings, using *their* language. If we have a real
> example (should be more!) on how data integration solves a specific
> development problem within an entreprise (even if this is not their
> specific area of business) *then* they will listen. Something that I
> have been telling for some time: we desperately need *those* use cases.
> With all my due respect (I mean it!) for the blogosphere, the Web2.0
> buzz, etc, I am not sure *that* will be the decisive factor to look at
> the Semantic Web for the Boeings, Mercks, DaimlerBenz-s, Shells, etc, of
> this World...
>
> Just my to pence. Well: the reason why I think it is imperative for the
> SWEO group to collect those use cases. In my view, we desperately need
> those industrial use cases (personally, I believe, more than anything
> else!) to help the Semantic Web moving forward.
>
> Apologies for the long mail...
>
> Cheers
>
> Ivan
>
> P.S. The HCLS IG will produce some documents soon on how these
> technologies can be used in their settings, by the way. And a bigger
> demo should be around by May. Is the correct, Susie?
>
>
>
>   
Ivan,

As you may recall, I have tried to have us develop our messaging from a 
base such that we split the usage scenarios, FAQ, and other collateral 
along the following lines:

1. Web Community
2. Enterprise Community
3. DBMS Technology vendors

The SIOC demos and commentary are about the Web (in particular the 
Blogosphere).  But it can be used for Enterprise customers if put in 
their context using appropriate parlance. The problem I see re. 
contributions to this discourse via SWEO is: how many members of SWEO 
currently fit the Enterprise Developer/Architect/Manager profile? I ask 
this question because if the readership of SWEO itself isn't conversant 
in "Enterprise Technology lingo" even the most basic contributions to 
such discourse will attract one of the following comments;

1. This is too technical (where this sometimes implies: I don't 
understand the lingo or I just don't like it, it's buzzword heavy etc.)
2. No! Product Promo etc..

The remainder of this mail is for SWEO readership in general (as opposed 
to Ivan specifically):

SIOC in the Enterprise is quite simple (IMHO), ditto the Semantic Web 
vision.

What does SIOC do for the Enterprise? It provides infrastructure for 
coherent implementation and usage of standards compliant Distributed 
Collaborative Applications (DCA) [1]. Every enterprise on this planet is 
struggling with the construction of cohesive and coherent infrastructure 
for collaboration. Just as the Web is making the world smaller, the 
behemoth enterprise of this planet are desperately trying to get 
smaller. Blogs, Wikis, Shared Bookmarks, Discussion Forums, Mailing 
Lists, IRC channels are all pieces of the solution puzzle that they are 
stifled in the enterprise for the following reasons:

1. Disparate Databases driving the platforms
2. Legacy platform lock-in problems at the DBMS, Operating Systems, and 
Development Environment levels

SIOC as an Ontology provides the glue for unshackling the Data in the 
aforementioned DCAs. Through SIOC they are able to map existing Data 
Sources to SIOC (and other Ontologies for that matter) using an array of 
emerging middleware solutions [2]. Of course, they don't have to use 
these middleware options, but this is where the basic cost-benefit 
analysis of the "build vs buy" comes in for the decision makers. Thus, 
when looking at the issue of collaboration, it is very clear to me what 
Semantic Web technologies offer to the Enterprises that are inherently 
heterogeneous at the DBMS, Operating System, and Development 
Environments levels. Even more so for someone who has worked on these 
Enterprise Integration problems since the late 80's (on both sides of 
the fence; I worked as an accountant for many years before crossing over 
to the vendor side of the fence).

Andersen Consulting have a nice presentation about enterprise DCA 
challenges titled: Visualizing Organizational Changes [3]. Although this 
particular case study deals with social networking (Graph Model etc..) 
and its impact on post merger relationships etc. You can also apply the 
same observations to the recent spate of A-List Bloggers that have 
departed their organizations to set up shop on their own. The point here 
is that organizations live and die by there ability to exploit 
Information. 

Another point, Lotus Notes is an example of first generation DCA 
technology.

SIOC is to the Semantic Web what RSS 2.0 was to the bootstrapping of the 
Blogosphere whch ultimately gave rise to the Web 2.0 movement. If you 
can juxtapose a SIOC button alongside every RSS 2.0 button in the 
Blogosphere and then query across the resulting data sources via SPARQL, 
you have the foundation for lots of practical Semantic Web 
demonstrations for Enterprise, Web, and DBMS communities. SIOC is 
flexible enough to pull along FOAF, SKOS, RSS 1.0, Annotea, AtomOWL and 
others endeavors in the RDF Instance Data for  Outreach and Education 
efforts. If these Ontologies weren't mutually inclusive what on earth 
are we doing? Or seeking to achieve? SIOC demos aren't aren't about SIOC 
vs FOAF vs Annotea vs anything else. They are simply about the quest for 
use cases that resonate with the target audience of SWEO: Decision Makers.

BTW - I am passionate about SIOC for the same reasons I am passionate 
about ODBC [4], neither is a creation of mine, I simply understand their 
intrinsic value and find myself lucky enough to be able to assemble the 
required resources to assist in exposing and sharing the value of said 
technologies with the broader community. 

Links:
1. http://www.cs.duke.edu/ari/cisi/relay/doc/paper.ps
2. http://esw.w3.org/topic/RdfAndSql
3. http://www.accenture.com/xdoc/en/AccentureSNA.swf (note this takes a 
while to load)

-- 


Regards,

Kingsley Idehen	      Weblog: http://www.openlinksw.com/blog/~kidehen
President & CEO 
OpenLink Software     Web: http://www.openlinksw.com

Received on Friday, 22 December 2006 14:38:12 UTC