[Top] [All Lists]

Re: [ontolog-forum] Webby objects

To: ontolog-forum@xxxxxxxxxxxxxxxx
From: John F Sowa <sowa@xxxxxxxxxxx>
Date: Sun, 18 Nov 2012 10:56:45 -0500
Message-id: <50A9053D.8080903@xxxxxxxxxxx>
Kingsley,    (01)

I've been highly critical of developments in the official W3C
recommendations that pretend that the rest of the world is going
to adopt their proposals as the universal direction for the future.    (02)

But I have been supportive of practical developments and commercial
products, such as yours, which recognize reality.  My criticisms
of the DAML project are not about the products that were produced
-- RDF, RDFS, OWL, and SPARQL -- but about the much more promising
products and directions that were killed, ignored, or deprecated.    (03)

>> But the SPARQL people keep saying that they "interoperate" with SQL
>> because they can convert an RDB to a bunch of triples.  The people
>> with triple stores who make a profit support SQL -- because that's
>> what the programmers ask for.    (04)

> No this isn't the claim. There are two orthogonal initiatives on this front:
> 1. R2RML -- a syntax for mapping relational database hosted data to RDF
>    model based entity relationship graphs
> 2. SPARQL -- an intensional query language for query RDF model based
>    entity relationship graphs (which includes transient or materialized
>    views of relational DBMS hosted data; ditto other RDF and non data 
>sources) .
> Today, as I've demonstrated many times [3], you can make transient and
> materialized RDF views over ODBC or JDBC accessible relational database
> hosted data [1]. In addition, you now have the ability to query the same
> data intensionally (via SPARQL) and extensionally (via SQL).    (05)

That's useful.  I have no complaints about developing useful tools.    (06)

But the distinction between intensional and extensional representations
and queries was a hot topic in the DB world in the 1970s.  The ANSI-
SPARC report in 1978 had three parts:    (07)

  1. Conceptual schema:  the ontology of a domain independent of any
     particular representation.  They used the word 'intensional', but
     they didn't use the word 'ontology', which didn't become popular
     until much later.    (08)

  2. Data schema:  the formats of the extensional representation of the
     data in tables (relational), networks (CODASYL DBTG), or trees with
     cross links (IBM's IMS, for example).  It was agreed that the same
     extensional information could be automatically translated to and
     from any of the data formats while preserving the intensions.    (09)

  3. Application schema:  the formats of application programs that
     accessed the data.  Any program with any preferred internal
     formats would be able to access information (extensional or
     intensional) in any data format with exactly the same requests.    (010)

This was published as an ANSI technical report in 1978, but it never
became a standard -- largely because of the vendors who had vested
interests in locking users into their proprietary formats.  The small
vendors wanted to make migration easy, but the biggest vendors (whom
I shall not name) most definitely did not want to make migration easy.    (011)

There were repeated attempts by ISO to develop a universal, logic-based
standard for the conceptual schema, but they also ended up as technical
reports -- the two most notable in 1987 and 1999.    (012)

My original enthusiasm for the Semantic Web is that it might finally
break through the stagnation by emphasizing logic and ontology rather
than proprietary data formats.    (013)

Tim B-L's proposal of 2000 looked like a new conceptual schema.  He
cited the latest research of the 1990s.  But the DAML project just
produced a warmed over YADM -- Yet Another Data Model.  There was not
a single innovation that used any ideas that had not been published
in the 1970s.    (014)

If anyone claims that DAML produced any innovations, please let me
know.  I believe that I can find a citation before 1980 for each.    (015)

John    (016)

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    (017)

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