W3C home > Mailing lists > Public > semantic-web@w3.org > January 2006

Re: [tangle] getting the semweb exactly wrong

From: Frank Manola <fmanola@acm.org>
Date: Tue, 03 Jan 2006 12:35:15 -0500
Message-ID: <43BAB5D3.5070005@acm.org>
To: Timothy Falconer <timothy@immuexa.com>
CC: semantic-web@w3.org

Timothy Falconer wrote:
> Blog post excerpt:
> "Reading such comments confounds me, since they've got it *exactly* 
> wrong. The Semantic Web approach is LOOSE, not normalized.
> ...


At the risk of being considered a "standards body formalist" (cf your 
blog), I have to wonder what you (and possibly others) mean when you use 
the term "normalized" in this context.  I'm assuming that this is a 
reference to "normalized" as used in relational data modeling.  However, 
if that's the case, from a relational data modeling perspective, RDF 
data is *highly* normalized: RDF essentially organizes data as binary 
relations (one per property) with surrogate keys (URIs), which is as 
normalized as you can get.  This high degree of normalization is one of 
the things that makes the data structure so flexible. RDF is looser than 
the relational model in some other respects, but they have nothing to do 
with normalization.  "Normalized" isn't properly the opposite of "loose" 
either;  you can have highly-normalized data structures that are 
centrally defined and controlled, and have lots of associated 
constraints (these can be appropriate on the Web too, but often they are 

I don't make this comment to be overly pedantic.  It's just that if 
you're going to try to explain the Semantic Web to "non-formalists" (in 
whatever discipline), it seems to me that a better strategy is to try to 
use non-technical descriptions that don't put people off, rather than to 
use technical terminology like "normalized" in an unfamiliar way. 
Besides, it might confuse database people who are (or at least ought to 
be) familiar with the concept of normalization.  The technical 
terminology was developed to make what, to those using it, are important 
distinctions, even if they aren't immediately important to 
non-specialists.  In addition, part of the job of explaining the 
Semantic Web to people is to explain why the "formalization stuff" so 
often associated with the Semantic Web is actually useful, even if it 
might not be apparent to start off with (step one in this process is 
often to explain what "semantic" has to do with it).

My $.02.

Received on Tuesday, 3 January 2006 17:33:46 UTC

This archive was generated by hypermail 2.4.0 : Tuesday, 5 July 2022 08:44:55 UTC