ontolog-forum
[Top] [All Lists]

Re: [ontolog-forum] Data, Silos, Interoperability, and Agility

To: ontolog-forum@xxxxxxxxxxxxxxxx
From: Kingsley Idehen <kidehen@xxxxxxxxxxxxxx>
Date: Wed, 25 Sep 2013 07:31:57 -0400
Message-id: <5242C9AD.3070101@xxxxxxxxxxxxxx>
On 9/25/13 2:30 AM, John F Sowa wrote:
> On 9/24/2013 5:25 PM, Kingsley Idehen wrote:
>> The expressive power of FOL needs to be meshed with the data being
>> processed.
> That statement requires many clarifications and qualifications.    (01)

I am implying that FOL should be an integral part of the underlying data 
representation. With that in place you can then use a query language to 
exploit expressiveness.    (02)

RDF does enable the incorporation of FOL into structured data 
representation.    (03)

SPARQL enables you query RDF model based structured data.    (04)

>
>> The data being processed is now disparately located, heterogeneously
>> shaped, voluminous, and volatile.
> That is why it is essential to design systems for which heterogeneity,
> diversity, and interoperability are of primary importance.    (05)

That's exactly what RDF based Linked Data is all about. The only problem 
is that RDF narratives haven't always made this most important virtue 
crystal clear.    (06)


>   Those were
> the goals of the DAML proposal of Feb 2000.
>
>> RDF based Linked Data enables the fusion of logic, data, and data access.
> No.  RDF was designed as a single data model that can only be useful
> if all the diversity is eliminated by mapping everything to it.    (07)

RDF simply adds IRIs (which can be HTTP URIs) and Predicate Logic to 
structured data representation. Historically, it's had a few awkwardly 
constructed draconian narratives, but nothing about it is inherently 
draconian.
> That is the antithesis of a multi-paradigm system that is designed
> from the beginning to support heterogeneity.    (08)

I disagree, profoundly. Modulo poor narratives, RDF is fundamentally 
about heterogeneity. I don't speculate, I can demonstrate these virtues 
with ease, as I do at every opportunity via links to real-world examples.    (09)

>
>> Using HTTP URIs as identifiers makes a big difference to many challenges
>> in this regard e.g., pointing to data across data spaces.
> Unique identifiers (AKA surrogates) have been used in DB systems for
> ages upon ages.    (010)

Of course they have. But none of what you've mentioned delivers the 
heterogeneity and data flow prowess of URIs. There's a really big 
difference here i.e., identifiers have existed for ever, but URIs build 
on the concept by not being silo vectors.    (011)

>   Various methods for globally identifying resources were
> proposed.  Among them were Unix-like file systems whose top level IDs
> would be a unique name for each separate file system (which could be
> accessed by multiple CPUs).  A version of this was adopted for Arpanet,
> which later was extended to the Internet.
>
> That was the basis for URLs.  URIs and IRIs are the next step.    (012)

Yes, not disputing the genealogy. My point is that URIs and IRIs are 
contemporary Identifiers for current data access, management, and 
integration challenges. They are essential to any practical 
data-de-silo-fication pursuit.    (013)

>    But the
> basis for the naming scheme was developed long before the SW, it was
> independent of the SW, it is certainly useful, but there are other
> naming schemes that could be just as useful or even more useful.    (014)

Please provide an example of an Identifier scheme that exists today, in 
broad use, that inherently enables data flow between disparate systems 
via reference and de-reference. In addition to that, how said identifier 
scheme is integrated into structured data representation that basically 
leverages FOL as the conceptual schema.    (015)

>
>> Missing link (order in my initial response is broken):
>>
>> [3] http://bit.ly/ZOCmaD -- Star Schema Benchmark Results (SPARQL vs SQL)
>> [4] http://bit.ly/10pvAbF -- Star Schema & The Cost of Freedom Blog Post.
> Those are not comparisons of SPARQL vs. SQL.  They are comparisons of
> certain kins of queries running on certain implementations of SPARQL
> and SQL.  With different queries for different purposes on different
> implementations, you can get radically different results.    (016)

Not really, those are useful queries that no SQL-only implementation 
that I know of -- as of the time of writing this post -- can execute in 
a practical manner i.e., I am yet to encounter a SQL query endpoint that 
can produce the same results in similar time. In fact, I am yet to 
actually find a live SQL endpoint on the Web (modulo ours) that allows 
anyone perform basic ad-hoc queries let alone the kind of queries I am 
demonstrating.    (017)

Does anyone know of a live SQL endpoint that allows ad-hoc queries?    (018)

SPARQL is a contemporary query language that addresses current issues. 
SQL is an old and inherently limited query language that simply doesn't 
stand up to contemporary data access, integration, and management 
challenges. I make this bold claim from the perspective of my own 
product (Virtuoso) which implements both query languages.    (019)

BTW -- I've just completed the migration of an entire system for 
handling online shopping, offers, and product cataloging. The original 
system was SQL RDBMS based, and its was an utter nightmare with regards 
to agility. The new system delivers all the agility (and more) that any 
modern company would seek circa. 2013. I am also going to have this 
entire endeavor documented and published for others to digest and analyze.    (020)


Kingsley
>
> John
>
>   
> _________________________________________________________________
> 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
>   
>
>    (021)


--     (022)

Regards,    (023)

Kingsley Idehen 
Founder & CEO
OpenLink Software
Company Web: http://www.openlinksw.com
Personal Weblog: http://www.openlinksw.com/blog/~kidehen
Twitter/Identi.ca handle: @kidehen
Google+ Profile: https://plus.google.com/112399767740508618350/about
LinkedIn Profile: http://www.linkedin.com/in/kidehen    (024)

Attachment: smime.p7s
Description: S/MIME Cryptographic Signature


_________________________________________________________________
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)

<Prev in Thread] Current Thread [Next in Thread>