W3C home > Mailing lists > Public > www-rdf-interest@w3.org > August 2004

Re: Reification - whats best practice?

From: Leo Sauermann <leo@gnowsis.com>
Date: Thu, 26 Aug 2004 09:56:26 +0200
Message-ID: <412D97AA.7080105@gnowsis.com>
To: Eric Jain <Eric.Jain@isb-sib.ch>
CC: Bob MacGregor <macgregor@isi.edu>, "'RDF interesting groupe'" <www-rdf-interest@w3.org>, danbri@w3.org
Hi Everybody,

First of all, thanks for the discussion
To sum it up:

Reification is a bad standard, so people flee from it into quads. There 
is no RDF/XML syntax that is usable with Reification.

*I think about joining DAWG :-)
*I want real life examples and real live RDF that fits real life use cases.

> Bob: As you noted, storage space explodes. 


Eric: >Only if implemented naively. Not much extra space is needed if 
reification is implemented by adding an identifier column to the 
statement table.

My email was about RDF/XML as storage space. No statement table. When I 
deploy real live applications, Idon't want to move my customers into 
installing some RDF storage system or cumbersome embedded database. 
RDF/XML should be fine. This is also for compability reasons. I am a 
scientist, it says so on my card, but our stuff should be useful and 
worthy (in terms of cash) for people and interoperable.


BTW Real life:
99% of all this Reification will be used to state things like:
- When was this triple added to the store? (date)
- By Whom ? (chown leo)
- who has access rights (chmod 777) to it?

this is metadata you will see in any store that is shared by one than 
more person (f.e. in a company environment, to use RDF as an 
organisational memory)

and these will be masses of triples and masses of reification/context.

I don't care if we call it context, reification, quadrupels. The 
implementation of these things look strikingly similiar. I had this RDF 
Gateway installed, it has quads f.e.



(Thats the sad thing: we built this stuff for selling Organisational 
Memories, Knowledge Management Software or Semantic Web Servers but all 
we talk about is proteins, museums and Wine collections. Instead we 
should talk about emails, documents, projects, departments, and people. 
Thats why we don'T sell yet.)



Es begab sich aber zu der Zeit 26.08.2004 08:40,  da Eric Jain schrieb:

>
> Bob MacGregor wrote:
>
>> It's rare that one statement captures what
>> you want to say, whether its a time-dependent statement or a 
>> probabillistic one, or whatever.
>
>
> I believe single statement reification is quite appropriate for 
> indicating why two resources are related.
>
> Example:
>
>   uniprot:P12345 rdf:type :Protein
>   uniprot:P12345 :organism taxonomy:9606
>   taxonomy:9606 :scientificName 'Homo sapiens'
>
> Now I want to indicate what the evidence is that the protein occurs in 
> the specified organism. How many statements does this affect? Exactly 1.
>
>
>> there were no "use cases"
>
>
> We have a lot of data that is backed by several sources.
>
> Consider the following example:
>
>   s1 p1 o1 : backed by a1 and a2
>   s1 p2 o2 : backed by a1 and a3
>
> If we were to use contexts for expressing this, there would have to be 
> three different contexts (for statements backed a1, a2 and a3), and 
> both statements would have to be duplicated into two different 
> contexts. Correct? I imagine this approach would bloat the data far 
> more than normal reification would...
>
> (If you still have doubts, consider this: My spell checker suggested 
> replacing 'reification' with 'deification' :-)
>
Received on Thursday, 26 August 2004 07:56:36 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 5 February 2014 07:14:57 UTC